WO1996005975A1 - Computer optimized adaptive suspension system and method improvements - Google Patents

Abstract

A controller (16) for controlling a damping system (12) is disclosed. The system (12) has at least two dampers (12a-12d) for damping between sprung and unsprung masses (7, 2) in at least one of compression and rebound directions. A sensor (190a-190d) generates position signals (17a-17d) representative of the displacement between the sprung and unsprung masses (7, 2). A regulator (40a-40b) responds to at least one of the independent compression and rebound control signals (25a-25d, 27a-27d) for adjusting, respectively, at least one of compression and rebound resisting forces of the dampers (12a-12d) between the masses (2, 7). The controller (16) includes a processor (15) that is responsive to signals representative of the position signals (17a-17d) for forming the compression and rebound control signals (25a-25d, 27a-27d) for the regulator (40a, 40b) as a function of motion between the masses (2, 7) or a motion of a vehicle (8) in which the dampers (12a-12d) are located.

Description

COMPUTER OPTIMIZED ADAPTIVE SUSPENSION

SYSTEM AND METHOD IMPROVEMENTS Cross Reference to Related Application

This is a continuation-in-part of application Serial No. 08/293,539, filed August 18, 1994, the subject matter of which is incorporated herein by reference.

Field of the Invention

The present invention relates to computer controlled vehicle suspension systems and methods, and more particularly to vehicle suspension systems and methods in which computer controlled damping forces in compression and rebound directions are used to optimize ride and handling characteristics of the vehicle.

Background of the Invention

Vehicle suspension systems have included shock absorbers, springs (coil, leaf, air or torsion bar), axle housings, torque arms, A-frames, anti-roll bars and stabilizers, and other elements. These components have been assembled in various combinations to produce the desired ride and handling characteristics for a particular vehicle. The characteristics of these devices often lead suspension system designers to compromise between various ride and handling characteristics and the design of such suspension systems. In a typical suspension system, changes in the spacing between the wheels/axles (unsprung mass) and the body /chassis (sprung mass) are cushioned by spring devices. Excess movement created by these spring devices must typically be controlled by damping devices.

The damping device (damper) is usually a velocity sensitive hydraulic system which uses hydraulic pressure to resist movement of a piston. Piston velocity is a direct function of the speed of suspension movement. The greater the piston velocity, the greater the force created in a direction opposite the movement. This force is created when the moving piston causes the hydraulic fluid, typically oil, to be forced through orifices and valves. The flow resistance encountered by the oil results in damping forces, both in compression and in rebound, and this damping acts to counter and dissipate the movement induced by the springs. Varying the fluid flow through the valves and orifices varies the forces acting against the spring-induced forces and, therefore, changes the ride and handling characteristics of the vehicle. The damping forces are passive or resistive forces such that the respective compression and rebound forces only have effect when the damper is being contracted and extended, respectively.

The typical spring force is proportional to spring deflection. The typical damping force is proportional to the velocity of the piston, that is, the velocity at which spring position changes. Modern hydraulic suspension systems include numerous complexities to account for this difference between the sources of these two forces, position and rate of change of position. For example, a 6-stage valve system, 3 in compression and 3 in rebound, is known.

Vehicle leveling systems for maintaining predetermined height between the sprung mass of the vehicle (frame and body) and the unsprung mass (wheels, drive train, front axle and rear axle) are known. Many of these systems utilize air springs wherein air is pumped into or out of such air springs to modify the position of the vehicle body relative to the wheels. Such systems are disclosed in U.S. Pat. Nos. 3,574,352, 3,584,893, 3,366,286, 3,830,138, 3,873,123, 4,017,099, 4,054,295, 4,076,275, 4,084,830, 4,162,083, 4,164,664, 4,105,216, 4,168,840, and 4,185,845. These systems, however, are not meant to actively adjust suspension system performance during vehicle travel to improve ride and handling characteristics.

Dynamic vehicle suspension systems have been developed. Systems for controlling the roll of a vehicle during a turn are disclosed in U.S. Pat. Nos. 2,967,062, 2,993,705, and 3,608,925. U.S. Pat Nos. 3,807,678 and 4,634,142 disclose active suspension systems which operate by changing the damping characteristics of the damping devices, based on a variety of discrete system states. These systems disclose utilizing a control system to respond to motion detected at the natural harmonic frequency of the suspension system to improve ride and handling characteristics. These control systems are not, however, optimum. These control systems may result in harsh handling characteristics or inadequate response to encountered road and driving conditions.

Other actively controlled vehicle suspension systems are disclosed in U.S. Pat. Nos. 2,247,749, 2,973,969, 3,124,368, 3,321,210, 3,502,347, 3,995,883, 4,065,154, and 4,215,403.

Summary of the Invention

There is accordingly provided in practice of the present invention a real-time computer control system for controlling one or more damping devices in a vehicle suspension system.

The damping devices damp movement between sprung and unsprung masses in both compression and rebound directions. The control system utilizes a sensor at each damping device for generating position signals indicating the relative position (displacement) of the sprung and unsprung masses as related by the extension of the damping device. Based upon these position signals, the control system iteratively determines the desired damping forces to ensure optimum ride and handling characteristics.

The desired compression and rebound resisting damping forces are determined, in part, by determining the amplitude of motion at the natural frequency of the sprung mass system by using an appropriate filter and then by applying proportional damping forces in the direction opposite to the displacement of the spring from its normal (equilibrium) position.

A constant force is approximately maintained on the vehicle (emulating a constant force spring), thus minimizing body motions. This force is calculated by multiplying the amplitude of motion at the sprung natural frequency by the actual amplitude signal defining the wheel to body displacement from its desired position, after performing an integration of the actual amplitude signal to filter out unwanted high frequency motions. Since the damping device is a dissipative device, in distinction to a true constant force spring, any natural frequency motions are automatically damped and an optimum control of the sprung natural frequency motion is achieved in a very simple and robust manner.

The amplitude of motion at the natural frequency of the sprung mass is determined through the utilization of a Discrete Fourier Transform (DFT). The DFT provides a selective response to a specific frequency such that a damping response can be provided for lower frequency inputs but higher frequency inputs can be ignored, thus weighing lower motion frequencies greater than higher motion frequencies. The DFT actually forms a narrow bandwidth filter that attenuates frequencies above and below the selected frequency that its parameters are set for. The DFT in the particular implementation detects the sprung and unsprung natural frequencies of the vehicle.

Advantageously, if the results of the DFT indicate a decreased amplitude of motion at the sprung natural frequency, then the amplitude of damping provided by the control system is also minimized. However, if the rate at which the value of the DFT amplitude of motion decreases is limited, then the rate at which the damping decreases is also slowed. This is useful so that as damped oscillations become very small, the damping does not cease too quickly (since the DFT output is also becoming very small).

In such a way, control forces are provided only when required and, even then, only at minimum levels required to control the sprung natural frequency motion. The amount of harshness encountered by vehicle passengers when the vehicle hits bumps at substantially random frequencies is drastically reduced. Thus, the response of the control system is a function of frequency, amplitude and time, thereby improving suspension system return to steady state conditions.

Further, initial integration of the actual amplitude signal prevents the unwanted application of damping forces in response to high frequency bumps when the movements caused by such bumps are superimposed on body movements at the sprung natural frequency. This provides an additional reduction in the harshness encountered by passengers. Particularly, an independent response in the compression and rebound directions such that the response is proportional to the extension of the damping device from its long-term average position provides improved ride and handling characteristics.

Application of a gain factor based on vehicular speed further improves overall suspension system response by reducing the amount of the damping forces as the vehicle slows down and smaller damping forces are required. This gain factor, however, is limited to a minimum value at very low vehicular speeds because the response to road perturbations encountered at those speeds requires a minimum level of damping (the damping cannot be reduced completely to zero for very low speed bumps such as those caused by parking lot speed bumps).

Brief Description of the Drawings

The features of the specific embodiment and mode of carrying out the invention are illustrated in the drawings, in which:

FIG. 1 is a diagram of a suspension system embodying the present invention;

FIG. 2 is a diagram of a vehicle having a computer controlled suspension system and embodying the present invention;

FIG. 3 is a block diagram of the computer control system and embodying the present invention;

FIG. 4 is a cross-sectional view of a damper unit suited for use in a suspension system and embodying the present invention;

FIG. 5 is a simplified definitional diagram of a damper;

FIG. 6 is a flow diagram of the computational process used by the computer control system of FIG. 2 and embodying the present invention;

FIGS. 7a-f are block diagrams of preconditioning calculations of the process of FIG. 6;

FIGS. 8a-g are block diagrams of the sprung natural frequency response calculations of the process of FIG. 6;

FIGS. 9a-c are block diagrams of the unsprung natural frequency response calculations of the process of FIG. 6;

FIG. 10 is a block diagram of the pumping down control response calculations of the process of FIG. 6; FIGS, l la-e are block diagrams of the roll control response calculations of the process of FIG. 6;

FIGS. 12a-e are block diagrams of the pitch control response calculations of the process of FIG. 6;

FIG. 13 is a block diagram of the topping and bottoming out control response calculations of the process of FIG. 6;

FIGS. 14a and b are block diagrams of the stored energy response calculations of the process of FIG. 6;

FIG. 15 is a block diagram of the float control calculations of the process of FIG. 6;

FIG. 16 is a block diagram of an unsprung natural frequency calculation of the process of FIG. 6;

FIG. 17 is a block diagram of a falling edge integrator calculation of the process of FIG. 6;

FIG. 18 is a block diagram of an alternate computer control system and embodying the present invention;

FIG. 19 is a block diagram of an alternate roll control response calculation of the process of FIG. 6; and

FIG. 20 is a block diagram of an alternate pitch control response calculation of the process of FIG. 6.

Detailed Description

Referring to FIG. 1, in accordance with the present invention, a wheel 2 engages a road surface 1 and is rotatably mounted on an axle 4 which extends from one end of a carrier 6. The other end of the carrier is pivotally mounted to a frame or chassis and body 7 of a vehicle 8. It is understood that a variety of wheel mounting configurations may be used. A suspension unit 10 is connected between the chassis and axle. The suspension unit combines a damping device 12 and a spring 14, which can be an air spring, coil spring, or similar device. The wheel, axle, and carrier comprise the unsprung mass of the vehicle, and the frame and body comprise the sprung mass of the vehicle. The damping forces FQ and FR in compression and rebound, respectively, of the damping device are varied by a controller 16, which incorporates the novel control features of the present invention, to optimize the ride and handling characteristics of the vehicle under a wide range of driving conditions and eliminate harshness associated with prior systems.

Referring now to the specific embodiment of FIG. 2, the vehicle 8 is shown comprising a steering mechanism 20, a hydraulic braking mechanism 22, four wheels, left and right front and left and right rear, 2a, 2b, 2c, 2d, respectively, suspended from the vehicle by suspension units as hereinabove described, and the controller 16 preferably a single controller associated with all four wheels. An exemplary vehicle is a 6000 lb. truck for which a specific embodiment of a suspension system is described. A damping device 12a-d, such as the damping device of FIG. 4, and a spring 14a-d couple each wheel, 2a-d respectively, to the body of the vehicle. The controller 16 receives position data (P_A/D) ^via lines 17a-d from position sensors 190a-d on each one of the damping devices 12a-d, respectively, indicating me position of the extensible member of each damping device. The subscript "_A/D" indicates that the particular variable reflects raw, real-time data which is the output of an analog to digital converter associated with a sensor device (not shown) in each one of the sensors 190a-d. Other sensing mechanisms that indicate wheel to body displacement that provide equivalent data could be used but may not be as cost effective.

The controller 16 also receives brake pressure (ρ_A/D) via lnie 21, steering angle (θ_A/D) via line 19, and vehicle speed (S _A/D) via line 23 from sensors 28, 30, and 32, respectively. The controller, in a continuous iterative manner, uses this data to determine the desired damping forces in the compression and rebound directions for application by the damping devices at each wheel. In this iterative manner, the controller uses such determinations to generate damper control signals for each damping device, to effect the desired damping forces in the compression and rebound directions, respectively. The signals are delivered to the dampers 12a-d via lines 25a-d and 27a-d, respectively, for compression and rebound. For the described embodiment, this continuous iterative process repeats at a fixed cycle or interval τ (referred to hereinafter as "cycle" or "iteration"), which may be, by way of example, every 5 milliseconds, based upon a computer driven interrupt timer in the controller, although one skilled in the art would realize that a variety of timing mechanisms and frequencies are possible. Because of this iterative process, where particularly useful for an understanding of the invention, the further subscript " _{n- 1} " indicates the iteration immediately preceding the present iteration denoted by "n".

Exemplary controller 16 is a microcomputer having a microprocessor such as a Motorola model 68332 processor with 16 MHz/16 bit performance as well as appropriate ROM, RAM, power supply and related circuity. An exemplary brake pressure sensor 28 is an EPNM-38-1000G by Entran. The various sensors with analog outputs preferably have analog output signals which are converted to digital form for the controller to perform the required calculations. A series of Analog-to-Digital (A/D) converters are provided for this purpose as described above. Analog-to-digital (A/D) converters are associated with each sensor. Such converters can, optionally, be included in the computer which serves as the controller. In the described exemplary embodiment, the output of the brake pressure sensor is monitored by a unipolar 8 bit minimum resolution A/D converter, providing a resolution of 4 PSI per least significant bit (LSB).

Exemplary steering angle sensor 30 can be in the form of a linear potentiometer attached to a control arm or other member of the vehicle steering gear. A suitable potentiometer is the "M-Series" by Maurey Instruments Corp. of Chicago, Illinois. The output of the potentiometer is monitored by a bipolar 8 bit minimum resolution A/D converter 29 (± 128 levels), providing a resolution per LSB of 0.5 degrees of wheel turn. Alternatively, an optical encoding system can be used wherein, for example, an optical sensor detects an array of lines or other markings on a steering column for detecting rotation of the column.

Exemplary speed sensor 32 is a Hall-effect sensor which detects a magnet connected to a drive shaft or other powertrain component. Many modern vehicles include some form of speed sensor for producing digital speed information, often associated with a vehicle transmission or with an anti-lock braking system (in which case a sensor is present for each wheel). The A/D converter 33 is similar to converter 29 with a resolution of 1 mph/LSB.

As is described hereinbelow, the position sensors 190a-d (shown in FIG. 2) can each comprise a linear position transducer connected to a hydraulic reservoir of the damper for measuring fluid displaced by the damper shaft as the damper is compressed. A suitable transducer is the Series 240 (such as model 0243-0000) by Trans-Tek Inc. of Ellington, Connecticut. The output of the potentiometer is monitored by bipolar 12 bit A/D converters 26a-d (shown in FIG. 2) (±2048 levels), thereby providing a resolution of 2/1000 inch per LSB for a sensor stroke range of ± 1 inch. The actual resolution used is 1/100 inch per LSB as referenced at the wheel, which allows for a larger range of input to cover expansion of the oil in the damper over wide automotive temperature ranges (typically -40° to +250°F). This expansion can be equivalent to 3-5 times the actual displacement of the damper shaft. This temperature range does not present a significant problem, however, since the resolution of the input P_A/D allows sufficient range and the long term integration and adjustment used to calculate an adjusted position P compensates for the offset.

Each of dampers 12a-d has a corresponding signal converter 18a-d, respectively, for converting digital compression control and rebound control signals from computer or processor 15 to pulse width modulated (PWM) signals for control of regulators 40a and 40b in each damper and for converting analog position signals from the position sensors 190a-d in dampers 12a-d, respectively, to digital position signals for the computer or processor 15. Each of signal converters 18a-d is essentially identical and, therefore, only signal converter 18a will be described by way of example, it being understood that the others are alike. Digital compression control and rebound control signals are converted from digital to analog and then to pulse modulated signals for compression and rebound regulators 40a and 40b; respectively, by digital to analog converters 52 and pulse width modulator circuits 24. Analog position signals from position sensor 190a is converted to a digital signal by analog to digital converter 51.

Referring to FIG. 3, the controller 16 receives data from sensors 32, 30, and 28, respectively, corresponding to the vehicle speed (S_A/D), brake pressure (P_A/D), and steering angle (θ_A/D) of the vehicle. Additionally, the controller receives from the position sensors 190a-d for each wheel 2a-d, respectively, through analog-to-digital converters 26a-d, the positions (P_A/D) 17a-d or extensions of the extensible member of the damping devices 12a-d at that wheel, such position or extension indicating the wheel to body displacement. The controller 16 utilizes these data inputs to determine the appropriate and desired damper forces in the compression and rebound directions for each wheel.

The controller 16 has input ports for receiving signals from sensors, output ports for providing control signals, and a central processor or computer 15. The computer includes the standard elements such as CPU, RAM, ROM and the like for processing the input signals to provide the desired outputs. The input and output ports can be formed as separate receiver and output portions of the controller or can be appropriately integrated within the computer. Similarly, the output signals from the controller to the pressure regulators in each damper 12a-d must be converted from digital form to a form appropriate for their correct operation. Preferred regulators 40a-b as shown in FIG. 4 are controlled by the amount of electrical current received. Although a controlled current can be achieved by providing the regulators with an adjustable current or voltage source, this type of analog circuitry wastes a considerable amount of power as heat. In the preferred embodiment, the output signals of the controller are received by pulse- width modulator (PWM) converter circuitry 24 (FIG. 3) which receive converted control signal from digital-to-analog converters and, responsive thereto provides a Pulse-Width-Modulated (PWM) signal that is then applied to each of regulators 40a-d for controlling that regulator for adjusting the force applied by the damper in compression or rebound, respectively. The PWM converter circuity 24 uses little power by pulsing the voltage to the regulators ON and OFF with varying duty cycles to achieve varying currents. The power is saved because the PWM voltage switching mechanism requires a minimum of power. All of the above circuit techniques are well known in the art. It should also be appreciated that other forms of analog (or digital) devices could be provided for the regulators, depending upon the particular regulator implementation used.

A specific embodiment of the damping device 12 is shown in FIG. 4, the damping device being configured for use in an automotive or similar vehicle shock absorbing system. The structure and operation of the damping device is discussed in further detail in our co- pending application Ser. No. 08/272,208, filed July 8, 1994 and international application number US 95/08550, filed

July 7, 1995, the disclosure of both of which are incorporated herein by reference. Briefly, the damping device 12 comprises an extensible member or shaft 160 within a body or case formed of concentric inner and outer sleeves 150 and 152. As the suspension respectively compresses and rebounds, the extensible member axially reciprocates inside the case forcing hydraulic fluid, such as oil, through a first (compression) relief valve 40a and a second (rebound) relief valve 40b, which relief valves regulate pressure in the compression and rebound directions, respectively. The blow-off pressures of these valves may be adjusted by the controller through the application of various voltages to solenoids 116a and 116b, thereby altering the compression and rebound damping forces applied by the damping device between the sprung and unsprung masses. A port 180 is formed in the outer sleeve of the second valve 40b. A hose 182 connects the port to a cylinder 184 in which rides a piston or extensible member 186.

To serve as the position sensor, a linear position transducer 190 is connected to the piston 186 for measuring its displacement caused by the displacement of fluid within the damper due to the varying length of the portion of the extensible member within the damper body. The signal indicative of the position (P_A/D) of the extensible member is provided by the transducer 190 to the control system via wiring pairs 192 which can coincide with wiring 17a-d for the four wheels, only one wire being shown for each of 17a-d for simplicity.

An exemplary solenoid 116a and 116b is the model PS-16 proportional solenoid by Elwood of Milwaukee, Wisconsin. An 8 bit digital-to-analog converter (not shown in FIG. 4) (256 levels) and PWM driver (shown in FIG. 3) is associated with the solenoid for providing necessary analog voltage based on the digital control signals from the controller. Such a converter may, optionally, be included in the microcomputer which serves as the controller 16.

FIG. 5 shows the damping device 12 in its maximally compressed (retracted) state 60 (solid lines). The maximally extended state 62 and long term average or equilibrium (normal) state 61 are shown in broken lines. The received position signal P_A/D is also pictorially shown, indicative of the displacement of the extensible member relative to a zero position 63 of the sensor. The signal is positive while the suspension is compressed and negative while the suspension is extended. Although it is desirable that the zero position of the sensor correspond to the equilibrium position of the extensible member, hardware tolerances and drift render this hard to achieve. Accordingly, as is described below, a baseline position signal P_LT1 is determined as a long term average, using a long term integration. The adjusted position signal P (relative to equilibrium) is then P_A/D - LT1. Similar adjustments are used for the outputs of the other sensors.

FIG. 6 is a schematic block diagram which shows the general steps of the computational process utilized by the computer controller to execute the control system of the present invention. The sequence is controlled by a computer program, preferably firmware in the computer controller. Calculations as discussed herein are performed by the computer under program control.

First, preconditioning (baseline) calculations (Block 522), more completely described hereinbelow in FIGS. 7a-f, are performed by the controller to provide the vehicle adjusted (baseline) speed S (FIG. 7e), steering angle θ (FIG. 7a), brake fluid pressure p (FIG. IT), position P (FIGS. 7b-c) and velocity V_p (FIG. 7d) of the extensible member of each damping device. Second, the various compression and rebound responses for each damper are calculated for a plurality of response modes and are more completely described hereinbelow in FIGS. 8-20. The response modes include a Sprung Natural Frequency response F_CSNF and F_RSNF (Block 524), an Unsprung Natural Frequency response F_CUNF and F_RUNF (Block 526), a Pumping Down response F_CPD and F_RPD (Block 528), Pitch and Roll responses F_CPH and F_RPH an F _CR anc F_RR (Blocks 532 and 533), a Topping Out and Bottoming Out response (Block 534), a Stored Energy response F_CSE and F_RSE (Block 536), and a Float response F_CFL and F_RFL (Block 538). Finally, the responses are summed (Block 540) to yield the overall compression and rebound responses F_C and F_R for each damper. The response modes are now defined.

The term "sprung natural frequency" (SNF) identifies the frequency at which the sprung mass (the frame and body) tends to oscillate on the springs. This can be determined empirically for a particular vehicle, but for a typical car or light truck is on the order of a 1.5 Hz frequency. The SNF response calculations (Block 524) determine the damping forces in the compression and rebound directions which will damp the vehicle motion due to spring oscillation. As more completely described hereinbelow in FIGS. 8a-g, this is done, in part, by determining the amplitude of body to wheel motion, as represented by the amplitude of motion of the extensible member at the sprung natural frequency, and by determining the current position of the extensible member. This information is used to provide a force proportional to and counteracting against the spring force inducing oscillation.

The term "unsprung natural frequency" (UNF) identifies the frequency at which the unsprung mass (the wheels, axles and carriers) tends to oscillate on the springs between the sprung mass and the road surface. As with the SNF, this can be determined empirically, but for a typical car or light truck is on the order of a 10-15 Hz frequency. The UNF response calculations (Block 526) determine the damping forces in the compression and rebound directions which will damp the wheel motion due to spring oscillation. As more completely described hereinbelow in FIGS. 9a-c, this is done, in part, by determining the amplitude of body to wheel motion, as represented by the position of the extensible member at the unsprung natural frequency, and providing a counterforce to damp this movement.

The term "pumping down" (PD) identifies the situation wherein the shock absorber compression forces during rapid wheel movements are less than the rebound forces, such that over an interval of time the net or total resulting force on the chassis or body is predominantly downward, pulling the chassis lower to the ground. Simply put, when the rebound force exceeds the compression force of the damper, the rebound force hinders the return of the springs from a compressed state upon oscillation of the vehicle and thus decreasing the average extension of the damper during the oscillation. As more completely described hereinbelow in FIG. 10, the Pumping Down response calculations (Block 528) determine when the chassis is being pulled downward and decreases the rebound resisting force to counteract the pumping down effect caused by otherwise excessive rebound forces resisting return of the compressed spring.

The term "bottoming out" identifies the condition wherein a road perturbation, such as a bump or other influence, causes the wheel and axle to reach the upper limit of the dynamic range of travel of suspension, that is, the suspension system is compressed or retracted to its maximum limit. Similarly, the term "topping out" refers to the condition where a road perturbation, such as a hole or other influence, causes the wheel and axle to reach the lower limit of the dynamic range of travel of the suspension, that is, the suspension system is extended to its maximum limit. The Bottoming and Topping Out (B/T) response calculations (Block 534) determine the damping forces required to avoid the occurrence of the aforementioned situations. As more completely described hereinbelow in FIG. 13, this is done, in part, by using the current position of the extensible member, the rate of change and the remaining travel from that position to determine the desired compression and rebound forces to avoid bottoming or topping out.

The terms "roll" and "pitch" refer to rotational body movement, about respective longitudinal and transverse axes of the vehicle, due to respectively lateral (transverse) and longitudinal accelerations caused by turning (roll) and braking or hard accelerating (pitch). Braking tends to force down the front of the vehicle while raising the rear of the vehicle

(compressing the suspension at the front wheels and extending the suspension at the rear). Accelerating causes the opposite to occur. Turning causes similar movements, but with respect to the left and right sides of the vehicle (a left turn compresses the suspension on the right side and extends it on the left side and a right turn causes the opposite). The Pitch and

Roll response calculations (Blocks 532 and 533) determine the damping forces required to counteract this body movement. As more completely described hereinbelow in FIGS, 11a-e, this is done, in part, by using sensors which indicate that at least one of roll or pitch is being induced and is otherwise imminent. Counteracting forces are then provided with the dampers to attempt to prevent corresponding roll or pitch motions, including feedback stabilization. The positions of the extensible members about the transverse and longitudinal body axes are compared to determine the desired compression and rebound damping forces. For example, to determine the desired compression and rebound forces for the damper associated with the left front wheel, the left front wheel extensible member position is compared to the right front wheel extensible member position for roll response and to the left rear wheel extensible member position for pitch response. An alternate pitch calculation comprises a comparison of diagonally opposite wheels.

The term "stored energy" (SE) identifies body movement along a diagonal axes of the body due to non-symmetrical road perturbations, that is, those which affect only a single wheel or combination of wheels at one end or side of the vehicle. The Stored Energy response calculation (Block 536) determines the damping forces required to counteract this body movement. As more completely described hereinbelow, this is done, in part, by comparing the positions of the extensible members of diagonally placed (opposite) wheels to determine the desired compression and rebound damping. For example, to determine the desired compression and rebound forces for the left front wheel, the left front wheel extensible member position is compared to the right rear wheel extensible member position. If the left front wheel hits a large bump, the right rear damper applies a counteracting compression force to prevent the vehicle from rotating downward in the rear about its center of roll in response to the upward force exerting on the front (left front wheel) via the compressed left front spring.

The term "float" identifies the movement which the sprung mass (the frame and body) tends to experience in the form of small rotations in roll and pitch and vertical motion. These motions and rotations, although uncomfortable to passengers during extended driving, are often too small for previous algorithms to detect and control. The specialized Float response calculations (Block 538) determine the damping forces which damp these subtle motions. As more completely described hereinbelow in FIG. 14, this is done, in part, by determining the amplitude of vertical, roll and pitch body motions, and using this information to provide a small "Coulomb" (constant) force that damps the oscillations. For each mode, the responses are calculated for both the compression and rebound directions. The results for each response are then summed (Block 540) to provide the total desired compression and rebound damping forces. These damping forces are translated into signals sent from the control unit to the regulator based on the characteristics of the damping device. Such translation can occur via a predefined look-up table, fixed equations or similar means. The preconditioning calculations (Block 522) will now be described.

The preconditioning calculations for the steering angle provide an adjusted and clipped steering angle. Although the steering angle position sensor is calibrated to provide a zero reading for straight ahead motion, the data received from the sensor is adjusted to reduce error due to miscalibration or sensor drift. The steering angle is also set to zero at low angles to additionally reduce error because the roll forces generated by small angle turns do not materially affect the ride and handling characteristics of the vehicle.

The detailed calculations to perform the adjustment and clipping of steering angle θ are shown in FIG. 7a. First, the angle θ_A/D is received (Block 702) and the counter Cθ is compared to N_θ (Diamond 704). If counter C_θ is greater than or equal to N_θ, C_θ is reset (Block 708). If C_θ is less than N _θ, C_θ is increased by adding 1 (Block 706) and the averageθ_LTI remains unchanged from the previous long-term integrated steering angle θ_LTIn-1 (Block 707). In the described embodiment, N_θ corresponds to a selected number of cycles equalling a set period of 0.64 seconds. Next, the magnitude of the received steering θ_A/D is compared (Diamond 710) to the magnitude of a constant angle θ_LIM' 5 degrees in the described embodiment. If θ_A/D has a magnitude greater than or equal to the magnitude of θ_{L IM'} θ_LTI remains unchanged (Block 712). The comparison of Block 710 is performed to avoid attempting to adjust for errors in the calibration of the sensor during high magnitude turns. If θ_A/D has a magnitude less than the magnitude of θ_LIM' θ_A/D iS compared to the previous long-term integrated steering angle θ_LTIn-1 (Diamond 714). If θ_A/D is greater than θ_{LTin-1 '}θ_LTI is increased by a small amount Δ_θ (Block 716), 0.5 degrees in the described embodiment. If θ_A/D is less than θ_{LTIn-1 '} θ_LTI is decreased by the small amount Δ_θ (Block 718). If θ_{A/ D} is equal to _{LTIn-1 '} θ_LTI is not changed (Block 707). The adjusted steering angle θ is then calculated by subtracting the long-term integrated steering angle θ_LTI from the received steering angle θ_A/D (Block 720). θ is further set to zero (Block 724) if the magnitude of θ is less than a minimum value θ_CLIP (Diamond 722), by way of example, 1 degree in the described embodiment.

The preconditioning calculations for the position of the extensible member 160 or 186 (FIG. 4) in each damper 12a-d as indicated by the signals on lines 17a-d provide an adjusted position as described below. FIG. 7b shows the preconditioning calculations for the long term integrated equilibrium position P_LTI of each of the extensible members. The position P_A/D is received (Block 726). The previously calculated θ (FIG. 7a) is compared to zero (Diamond 727) to prevent the long term integrated position P_LTI from changing when it should not. For example, cornering results in temporary changes in position P_A/D that do not represent the average normal ride position desired. A counter (Diamond 728 and Blocks 729 and 730) is used similar to that used in the preconditioning calculations for 0 in FIG. 7a. If the value of steering angle θ is not zero, the integrated long-term position P_LTI is not changed (Block 732). If the value of θ is zero, then at a set period, 250ms in the described embodiment which corresponds to a selected number of cycles N_P of the system, the received extensible member position P_A/D is compared (Diamond 734) with the previous long-term average position value P_LTIn-1. If the received extensible member position P_A/D is greater than P_LTIn-1, P_LTI is increased by adding a small amount Δ_p, such as 1/100 inch in the described embodiment, to the previous value (Block 736). If the received extensible member position P_A/D is less than P_{LTIn-1 '} P_LTI ιs decreased by subtracting the small amount Δ_θ from the previous value (Block 738). If P_A/D is equal to P_{LTIn-1 '} P_LTI ιs not changed (Block 732).

As shown in FIG. 7c, every cycle, the adjusted position relative to equilibrium of the extensible member P is calculated. P_A/D is received (Block 739) and the long term integrated equilibrium position P_LTI is subtracted from P_A/D (Block 740). Although N_P can be 1, if the integrated rate must be slow, the addition or subtraction of the smallest amount during each cycle might still be too fast. In this case, a counter is implemented so that a small amount is added after a repeated interval of a larger number of cycles Np. For example, an integration of P_LTI in the described embodiment adds or subtracts 1/100 inch every 50 cycles, each cycle being 5ms, thus having a 250ms delay between increments.

FIG. 7d shows, for each extensible member 17a-d, the calculation of damper velocity V_P. The current received position P_A/D (Block 755) is received. The previously received extensible member position P_A/Dn-1 is subtracted from the current received position P_A/D and the difference is divided by the cycle т (Block 756). The rate of change V_P of the position of the extensible member is calculated as an aid to the hereinbelow described Bottoming and Topping Out response (Block 534).

FIG. 7e shows the determination of the integrated speed S of the vehicle which provides an improved controller response. The speed S_A/D is received (Block 757). At an interval of N_S cycles, 100 in the particular implementation, as determined by a counter (Diamond 758 and Blocks 759 and 760), the received vehicle speed S_A/D is compared to the integrated speed S_n-1 (Diamond 761). If the received vehicle speed S_A/D is greater than the previous integrated speed S_n-1, then the integrated speed S is increased by a small amount Δ_S (Block 762). If the received speed S_A/D is less than the previous integrated speed S_n-1, then the integrated speed S_n-1 is decreased by a small amount Δ_S (Block 764). If S_A/D is equal to S_n-1, then S is not changed (Block 766). In the described embodiment, by way of example, the small amount Δ_S is 1 MPH every 500ms to achieve the correct integrated rate. The preconditioning calculations for the brake pressure ρ are similar to the preconditioning calculations for the steering angle θ, however, brake pressure is assumed to be positive. In embodiments wherein an indication of acceleration that can be used to predict both dive and squat is present, such as with an accelerometer, the parameter for brake pressure would have both positive and negative values. As shown in FIG. 7f, the pressure

P_A/D is received (Block 770) at a set period, 250ms in the described embodiment as determined by a counter (Diamond 772 and Blocks 774 and 776), which corresponds to a selected number of cycles N_ρ of the system equal to 50. The magnitude of the received brake pressure ρ_A/D is compared (Diamond 778) with the magnitude of a constant pressure ρLIM' 25 psi in the described embodiment. If ρ_A/D has a magnitude greater than or equal to the magnitude of the constant pressure ρ_LIM' then the long-term integrated brake pressure ρ_LTI will not be changed (Block 780). If ρ_A/D has a magnitude less than the magnitude of the constant pressure P_LIM' the n ρ_A/D is compared with the previous long-term integrated brake pressure ρ_{LTI n-1} (Diamond 782). If ρ_A/D is greater than ρ_{LTIn- 1 '} ρ_LTI is increased by a small amount Δ_ρ (Block 784), 1 psi in the described embodiment. If ρ_A/D is less thanρ_{LTIn-1 '} ρ_LTI is decreased by the small amount Δ_ρ (Block 786). If ρ_A/D is equal to ρ_{LTIn- 1} then ρ_LTI is not changed (Block 775). The adjusted brake pressure ρ is then calculated by subtracting the long-term integrated brake pressure ρ_LTI from the received brake pressureρ_A/D (Block 790). ρ is further set to zero (Block 794) if the magnitude of ρ is less than a minimum value ρ_CLIP (Diamond 792), by way of example, 10 psi in the described embodiment. The response modes will now be described.

FIG. 8a shows the steps of calculating the Sprung Natural Frequency response (Block 524) which are subsequently described in detail in FIGS. 8b-g. An integration or filtering of the extensible member position relative to equilibrium, P, is performed (Block 802), resulting in an integrated position P_SNF of the extensible member. Such an integration can be either a linear or non-linear integration, with the described embodiment, by way of example, using a non-linear integrator to increase controller response to large perturbations. This integration allows movement at the sprung natural frequency to be detected, while filtering out unwanted movements at higher frequencies. Any common low pass or bandpass filter technique can be used to detect the sprung natural frequency. In the described embodiment, a Discrete Fourier Transform is performed on the integrated position P_SNF (Block 804), resulting in an intermediate parameter A_DFTSNF , and thereby providing the amplitude of the motion of the extensible member at a selected frequency, that is, the sprung natural frequency. The intermediate parameter A_DFTSNF is modified by being multiplied by a vehicle dependent gain factor and a vehicle speed dependent gain factor to increase the controller response at higher speeds, which is also clipped or limited to a maximum value (Block 806), yielding A_SNF. In turn, A_SNF is smoothed (Block 808) to yield an integrated parameter A_INTSNF, an operation which is later described in detail, when the magnitude of the intermediate parameter A_DFTSNF is decreasing. The smoothing results in increased controller response as motion is damped. The integrated parameter A_INTSNF is then multiplied by both the absolute value of P_SNF and a vehicle dependent gain factor (Block 810), resulting in an intermediate parameter proportional to the amplitude of motion of the extensible member at a selected frequency and the current displacement of the extensible member. This parameter is then used to provide forces which are proportional to and in an opposite direction to the spring force F_RSNF and F_CSNF, respectively. Finally, if the Roll and Pitch algorithms are already providing forces, only the amount exceeding such already- existing forces is provided.

FIG. 8b shows the details of the non-linear integration (Block 802) of the position of the extensible member. Every iteration, the position of the extensible member P is compared to the value of the previously calculated integrated position value P_SNFn-1 (Diamond 820). If P is greater than P_SNFn-1 then an integrated value is added to P_SNFn-1 (Block 822). In the described embodiment, the integrated value being, by way of example, K_SNF | P | + 1, 0.1 being advantageously used for the value of K_SNF. If P is less than P_{SNFn-1 '} the integrated value is subtracted from P_SNFn-1 (Block 824). If P is equal to P_{SNF '} P_SNF is unchanged (Block 826). Thus:

P_SNF = P_SNFn-1 + K_SNF | P | + 1; P>P_SNFn-1

P_SNFn-1 - K_SNF | P | - 1; P< P_SNFn-1

P_SNFn-1 ; P=P_SNFn-1

The addition and subtraction of a " 1 " in the above equations compensate for round-off error in the algorithm. Since the position is being multiplied by 0.1, if the position falls below 10, the integration stops (0.1 x 9 = 0 integer). For the integrated value to return to 0 as required, a minimum integration amount is always provided.

FIG. 8c shows that a Discrete Fourier Transform (DFT) is performed (Block 804) on the integrated position value P_SNF resulting from the previous step of FIG. 8b. Performing such a calculation is a well-known mathematical technique. Other filter techniques would be equally acceptable. The described embodiment, by way of example, utilizes the values of P_SNF from the previous 135 iterations of the system. This number of points in the DFT is the fundamental frequency to be detected. 135 cycles at 5ms per cycle represents 675ms, which corresponds to approximately 1.5 Hz -- the desired frequency to detect in the described embodiment.

FIG. 8d shows the vehicle speed dependent gain factor K_SSNF being calculated as a function of the form S÷S_REFSNF, where S_REFSNF is a constant reference speed for which the speed dependent gain factor is 1 and S is the integrated speed of the vehicle. In the described embodiment, the value of this gain factor is not allowed to go below a minimum value, such value being determined by the value at an empirically determined specific vehicle speed S_MIN. In particular, if S > S_MIN (Diamond 844), K_SSNF is set to S÷S_REFSNF (Block 846), otherwise K_SSNF is set to S_MIN÷S_REFSNF (Block 848). In the described embodiment, the values are as follows:

CONSTANT VALUE

GAIN_SNF 1.0

S_REFSNF 50 MPH

S_MIN 30 MPH

In FIG. 8e, the result A_DFTSNF of performing the DFT is multiplied (Block 834) by the vehicle speed dependent gain factor K_SSNF (calculated above in FIG. 8d) and a constant vehicle dependent gain factor GAIN_SNF, which are both based on the characteristics of the vehicle. The result (Block 806) of this multiplication A_SNF is limited to a maximum value A_DFTSNFMAX, 1.0 inch in the described embodiment (Block 838), if the result of the multiplication would otherwise result in a greater value (Block 836).

FIG. 8f shows that A_SNF is smoothed when its value decreases, resulting in A_INTSNF (Block 808). This smoothing process helps maintain the damping level up as the motions are decreased so as to quickly and effectively stop the body motions. This is done every N_A cycles (1 in the described embodiment) determined by counter C_A (Diamond 850 and Blocks 852 and 854). The value of A_SNF is compared to the previous value of A_INTSNFn-1 (Diamond 856) and A_INTSNF is set to the value of A_SNF if A_SNF is greater than or equal to the previous value of A_INTSNF (Block 851). If A_SNF is less than the previous value of A_INTSNF, however, then a constant value Δ_A, 1 in the described embodiment, is subtracted from the previous value of A_INTSNF to form a new value of A_INTSNF (Block 858).

Referring to FIG. 8g, the integrated position | P_SNF | is multiplied by A_INTSNF and either a rebound constant (Block 860) or a compression constant (Block 862) to form the compression and rebound desired forces F_CSNF and F_RSNF (Block 810). Exemplary values for these constants are varied for the front and rear wheels of the vehicle and are listed below (wherein K_CSNFF indicates the front compression constant, K_RSNFR indicates the rear rebound constant, and so forth):

CONSTANT VALUE

K_CSNFF 0.50

K_RSNFF 0.60

K_CSNFR 0.65

K_RSNFR 1.00

The constants above are dependent on the resolution of the inputs and outputs of the overall system. In the described embodiment, the position P is represented in 1/100ths of an inch (P=100 indicates that the actual value of P is 1 inch). An output value of 1 results in an approximate damping force of 7 pounds. Therefore, as an example, if the above mentioned constant equals 1 and the vehicle is moving at a sprung natural frequency with an amplitude of 1 inch, the algorithm provides an output of 100 (1.0 constant X 100 DFT amplitude) which results in a damping force of approximately 700 pounds (100 X 7 pounds).

As these forces are somewhat duplicative of the roll and pitch forces, they are modified by the corresponding roll and pitch forces based on the integrated position P_SNF of the extensible member. If P_SNF is less than or equal to zero (Diamond 864), the compression desired force F_CSNF is reduced (Block 866) by the total of the compression and rebound forces determined by the Roll and Pitch responses, but not to less than 0 (Diamond 868 and Block 870) otherwise no rebound force is provided (Block 870). If P_SNF is greater than zero (Diamond 874), the rebound desired force F_RSNF is reduced (Block 876) by the total of the compression and rebound forces determined by the Roll and Pitch response, but not less than 0 (Diamond 878 and Block 880). Otherwise, no compression force is provided (Block 880).

U.S. patent 4,634,142 discloses that a filter can be used to detect when the sprung mass is starting to oscillate. The filter approach, however, was dismissed because the output of a filter has a delay and it was felt that the response would be too slow and the performance would be inadequate. Furthermore, the action that would be required to damp the mass if the filter was used is not described (that is, if there is an output indication of natural frequency motion, then how would that be utilized to damp the mass). It was found that the algorithm (and hardware) disclosed in the '142 patent controlled the mass very well. However, the ride was too harsh over bumps faster than the natural frequency of the masses because the displacement using signals only proportional to displacement of the masses damped all of the time, even for fast bumps above the natural frequency. As disclosed in the present application, modulating the displacement-only approach in the '141 patent with a filter as in FIG. 8c and 8g provides a ride performance that is improved significantly. Additional improvements include: 1) modulation of the forces by vehicle speed as in FIG. 8d; 2) pre-filtering of position to eliminate high frequency inputs to the DFT as in FIG. 8b; 3) limiting the maximum control forces as in FIG. 8e; and 4) a falling edge integrator to hold the damping force up as the oscillations damp out (and the output of the DFT falls) as in FIG. 8f.

The controller determines the response to wheel movement at the unsprung natural frequency in a manner similar to that of the sprung mass response. Referring now to FIG. 9a, as for the sprung natural frequency, a filter can detect the unsprung natural frequency. In the described embodiment, a DFT is performed (Block 902) on the adjusted extensible member position P to determine the amplitude of motion P_DFTUNF at the unsprung natural frequency. The adjusted extensible member position data is not filtered, as was done for the sprung natural frequency response, due to the negative effect it would have on determining the amplitude of the higher frequency unsprung motion. For a utility vehicle, a typical sprung natural frequency is on the order of 1 Hz, whereas the unsprung natural frequency is on the order of 10 Hz. The amplitude of motion P_DFTUNF can be optionally smoothed, in the same manner as was described for A_INTSNF above, when the current value of

P_DFTUNF has a smaller magnitude than the previous value. This is not required and not included in the described embodiment. Low amplitude motion at the unsprung natural frequency is optionally clipped and filtered (Block 908), however, such motion does not have appreciable effects on ride and handling characteristics but does reduce harshness on rough roads. The output of the DFT is then multiplied by a vehicle dependent gain factor (Block 910) to yield a desired rebound force.

The DFT, shown in FIG. 9b, is similar to that for the SNF mode but is performed on the previous 15 iterations. 15 cycles at 5ms per cycle represents 75ms which represents approximately 13 Hz -- the desired frequency to detect in the described embodiment.

An alternative and preferred approach to controlling the unsprung natural frequency is shown in FIG. 16. Just as for the previous UNF algorithm, a filter such as a DFT (Block 1610) determines if the unsprung mass is moving at its natural frequency, A_DFTUNF. This signal is then smoothed with a falling edge integrator (Block 1611) to hold the damping force up as the oscillations damp out. The integrator is shown in FIG. 17 and is essentially the same as the one for SNF as shown in FIG. 8f. The desired rebound damping force is then determined by multiplying the integrated DFT signal A_INTUNF, determined at Blocks 1675 and 1685 of FIG. 17, by the damper velocity V_p determined in FIG. 7d and a fixed gain parameter GAIN_UNF (Block 1620). The multiplication by the damper velocity ensures that the damping forces rise gradually as the wheel changes directions to significantly reduce harshness on bumps. To further limit harshness on very large bumps, a maximum limit is allowed for the force that is applied (Blocks 1625 and 1630). Finally, the force added to the suspension rebound control by the UNF algorithm is limited to that which exceeds forces that might already be present due to other algorithms (Blocks 1635, 1640 and 1645). This is because the UNF motion damping will occur when forces are applied by other algorithms so no further forces need be added except what is needed over and above what is already there. Preferably, N_UNF in FIG. 17 is 1 and K_UNF is 1.

The details of the clipping and filtering of low amplitude motion is shown in FIG. 9c. The amplitude of motion A_DFTUNF received from the DFT is compared (Diamond 920) with a first constant value, 10/100 inch by way of example in the described UNF_CLIP embodiment. If the amplitude of motion A_DFTUNF is greater than UNF_CLIP, it is compared (Diamond 924) with a second constant value which is the low amplitude exponential filter constant UNF_EXP. If the amplitude of motion is less than the exponential filter constant, the amplitude of motion is exponentially filtered to yield a filtering value K_UNFEXP by setting K_UNFEXp (Block 926) to a value equalling the amplitude of motion A_DFTUNF divided by the exponential filter constant UNF_EXP, quantity squared. Subsequently, as is described below, the amplitude A_DFTSNF is multiplied by the exponential filtering value K_UNFEXP. By way of example, the exponential filter constant UNF_EXP is 0.5 inch in the described embodiment.

If, after the comparison (Diamond 924) with the exponential filter constant the amplitude of motion is greater than the exponential filter constant, the filtering value is set to 1 so that subsequently the effective amplitude of motion remains unchanged (Block 928). The value K_UNFEXP, and thus the effective amplitude of motion, is set (Block 922) to zero if the amplitude of motion is less than the first constant value. The resulting filtered amplitude of motion (A_{DFTU NF} K_UNFEX ) is then multiplied by a vehicle dependent gain factor GAIN_UNF, 3.0 by way of example in the described embodiment, to determine the desired rebound damping force (Block 930). If the amplitude, however, is 0.25 inches, the gain factor is 1/4 (1/2 of the constant value of 0.5 inches results in a filter gain factor or exponential filter value of 1/4. Thus, a non-linear accelerated reduction in force occurs as the amplitude decreases below the constant limit). The compression force is optional, however, and in the described embodiment is zero. Conventional dampers typically divide the damping into 1/3 for compression and 2/3 for rebound.

U.S. Patent 4,634,142 discloses DFT used to detect the UNF frequency and then apply a compression force proportional to the DFT output. Later testing on a vehicle and conventional damper approaches showed that the force should be applied to rebound only and not compression. Therefore, the improved algorithm of FIG. 9 was developed that uses rebound only (Block 930) and also adds small amplitude clipping and exponential filtering to try and help reduce harshness as the UNF motion is damped as in FIG. 9c. Note, the novel use of modulating the force with the output of the DFT remains.

Subsequent testing on a vehicle has shown that the control algorithm may still be too harsh. Since the valves are constant force, the rebound force goes ON and OFF very abruptly as wheels move up and down (compression and rebound directions). The compliance built into the damper helps this some, but is limited. A much improved algorithm is disclosed herein that optimally handles the UNF motions with greatly reduced harshness, better even than can be done conventionally. The preferred algorithm is shown in FIG. 16 and includes: 1) the modulation of the force by the damper velocity (Block 1620), 2) the addition of a maximum limit on the force applied (Block 1625), 3) a falling edge integrator to hold the damping force up as the oscillations damp out (and the output of the DFT falls) (FIG. 16b), and 4) limiting the addition of UNF forces to the damper to those required that exceed forces that may already be there due to the other algorithms (i.e., the UNF damping will occur when forces are applied by other algorithms so it need not be over damped) (see Block 1640). These features provide a near optimum response for UNF. Pumping down control is added to the system response to provide for proper system operation, particularly to avoid introducing unbalanced forces primarily due to the unsprung frequency response into the system. The pumping down force is proportional to an integrated position P_INTPD of the extensible member, the vehicle dependent gain factor GAIN_PD being 0.7 in the described embodiment.

The integrated position P_INTPD forms an average value over a selected period of time. The integration provides an effective low pass filter whereby faster changing signals are attenuated over slower changing signals. The choice of integration constant Δ_PD and cycle constant N_PD determines at what frequency the attenuation starts to significantly take effect. This concept is well known in the art.

As shown in FIG. 10, the integrated pumping down position P_INTPD is determined by the current position P of the member at a predetermined interval, 250 milliseconds in the described embodiment, by using a counter having N_PD = 50 cycles (Diamond 1010 and Blocks 1012 and 1014). At each interval, the current extensible member position P is compared (Diamond 1018) to the previous integrated pumping down position P_INTPDn-1. If the current extensible member position is greater than the previous integrated pumping down position, the pumping down position is increased (Block 1024) by an integration constant Δ_PD, 1/100 inch in the described embodiment. If the current extensible member position is less than the integrated pumping down position, the integrated pumping down position is decreased (Block 1020) by the same integration constant Δ_PD. If not at the integration interval or P = p_{INTPDn-1 '} P_INTPD = P_INTPDn-1 (Block 1026).

The integrated pumping down position is then compared (Diamond 1030) to zero. If the integrated pumping down position is less than or equal to zero, the integrated pumping down position is multiplied by the vehicle dependent gain factor GAIN_PD (Block 1032), and a force is subtracted from the current compression force (F_RPD = 0 and F_CPD = P_INTPD

* GAIN_PD). If the pumping down position is greater than zero, the integrated pumping down position is multiplied (Block 1034) by the above gain factor, and force is subtracted from the current rebound force (F_RPD = -P_INTPD * GAIN_PD). Pumping down control is described in U.S. Pat. No. 4,634,142, and in James M. Hamilton, "Computer-Optimized Adaptive Suspension Technology (COAST)," IEEE Transactions on Industrial Electronics, Vol. IE-32, No. 4, November 1985, which also disclose damper and control hardware which can be utilized in a system incorporating principles of the present invention and the disclosures of which are incorporated herein by reference.

U.S. Patent No. 4,634,142 discloses that wheel position is integrated and if it is pumping down, additional force is applied to compression to balance the damping and limit the pumping down. The same basic approach but, rather than adding forces to balance the pumping down forces and thus adding more harshness, the preferred approach disclosed herein is to subtract from the rebound forces that are pumping the suspension down (see Blocks 1032, 1034) rather than adding to the compression forces.

FIGS, 11a-e show a computational process for determining controller roll response (Block 532). Each cycle, intermediate compression and rebound values are calculated using steering angle θ and vehicle speed S. A roll variable R, representing the force inducing a roll motion, is calculated by multiplying θ and S. An integrated roll variable, R_INT' indicating an actual amount of roll, is calculated using the difference in position of the extensible members of oppositely placed wheels. This integrated roll variable R_INT is then exponentially filtered to obtain actual roll R_EXP. These results are utilized in determining the desired compression and rebound roll forces.

Referring to FIG. 11a, the roll inducement R is determined by retrieving the vehicle speed S and steering angle 0 and multiplying the retrieved values (Block 1112). Intermediate compression and rebound values R_C and R_R, respectively, are computed by multiplying the roll inducement value R by vehicle dependent gain compression and rebound factors K_RC and K_RR, respectively (Block 1114) which in the described embodiment have values of 2.4.

Referring to FIG. 11b, a speed dependent roll constant K_SR is calculated by retrieving the vehicle speed S and comparing the speed to a minimum value S_MIN (Diamond 1120). If S is greater than S_MIN, K_SR is set to (S-S_MIN) ÷ (S_REFRP-S_MIN) (Block 1122). Otherwise, K_SR is set to 0 (Block 1124). In the described embodiment, S_REFRP is 65 miles per hour and S_MIN is 30 miles per hour.

Referring to FIG. lie, a differential position R_ACT is determined by retrieving the position P_LEFT of a left wneel extensible member and the position P_RIGHT of an opposite right wheel extensible member and subtracting P_RTGHT from P_LEFT (Block 1120). The resulting value, R_Acτ, is integrated (Block 1132), as has been previously described for other modes, yielding the integrated roll variable R_INT. the integration constant Δ_R used in the described embodiment is 5. This integrated roll variable is then exponentially filtered in a manner similar to that of the unsprung natural frequency amplitude to obtain K_REXP. In particular, if | R_INT | is less than the low amplitude filter constant R _EXP (Diamond 1134), K_REXP is set to the magnitude of | R_INT | (Block 1136). Otherwise, KR_EXP is set to the square of R_INT divided by R_EXP (Block 1138). In the described embodiment, R_EXP is 1/4 inch. This exponential filtering value is similar to that previously described for the pumping down response mode and forms an accelerated reduction in force as the amplitude decreases below the constant limit.

The desired rebound and compression forces are next computed as shown in FIG. 11d. These computations have been experimentally determined to provide improved ride and handling characteristics. The values K_SR, K_REXP, R_C and R_R are retrieved (not shown). A damping compression value COMP is calculated (Block 1142) by adding an optional roll- damping constant R_DAMP' which is 10 in the described embodiment, to the intermediate compression value R_C, and by multiplying that sum by the filtering roll value K_REXP, the speed-dependent roll gain factor K_SR and a vehicle dependent gain factor GAIN_RP. A damping rebound value REB is similarly calculated using the intermediate roll value R_R in place of R_C. The desired force to counteract the roll inducement COMPP is calculated (Block 1144) by doubling COMP and adding the resulting value to the product of the intermediate compression value R_C and the speed dependent roll gain factor K_SR. COMP is approximately 1/3 of COMPP and is applied in the opposite direction to that of COMPP to damp any overshoot as the vehicle attempts to correct itself. In addition, the constant value R_DAMP that is added provides roll and pitch control which is independent of any sensor indications that control is required. For example, when a vehicle is going straight ahead with a steering angle equalling zero, cross-winds causing roll body motions can be reduced by the position dependent control feature of the algorithm. Again, a desired compression force to counteract the roll motion REBB is calculated in the same manner as COMPP, but using R_R in place of R_C.

The forces COMP, COMPP, REB and REBB are applied to the respective left and right dampers depending upon the direction of induced roll and the direction of the integrated indication of actual roll (using its sign) as shown in FIG. 1 le. Specifically, the values R and R_lNT are retrieved (not shown) and if R is greater than or equal to 0 (Block 1152) (indicating that the vehicle is making a right turn), R_INT is compared to 0 (Diamond 1154) (indicating the vehicle is rolling left). If R_INT is greater than or equal to 0, for the left damper, the roll compression force F_CRLEFT is set to the value COMPP, the rebound force F_RRLEFT is set to COMP, the right compression force F_CRRIGHT is set to REB and the rebound force ^PRRRIGHT is set to REBB (Block 1156). If, however, R_INT is less than 0 (indicating the vehicle is rolling right), it is not necessary to counteract roll induced by me turn and, accordingly, the values are set as F_CRLEFT = COMP, F_RRLEFT = F_CRRIGHT = 0 and P_RRRIGHT = REB (Block 1158). Similarly, if R_INT is less than 0 (indicating the vehicle is turning left), R_INT is compared to 0 (Diamond 1160). If R_INT is less than or equal to 0, the vehicle is rolling right and F_CRLEFT = REB, F_RRLEFT = REBB, F_CRRIGHT = COMPP and F_{RRIR GHT} = COMP (Block 1162). Otherwise, the vehicle is rolling left and

F_CRLEFT = 0, F_RRLEFT = REB F_CRRIGHT = COMP and F_RRRIGHT = 0 (Block 1164).

The above forces are determined whenever the vehicle encounters a maneuver which causes roll or pitch, such as, cornering or braking. Counteracting forces are applied proportional to the amount of influence being applied to the vehicle, such as, the product of vehicle speed and turn angle. In addition, if any motion still occurs (indicating there is incorrect force), then the integrated actual indication of roll R_INT provides an error signal so that additional corrective forces are applied or removed in a stabilized feedback manner.

As shown in FIGS. 12a-e, the pitch response is determined in a similar manner with the following substitutions. The roll value R is replaced with a pitch value PH equal to the brake pressure p. This represents pitching forward as the vehicle brakes. Acceleration causes the vehicle to squat and any suitable sensor which indicates acceleration can be used as the equivalent of sensing positive brake pressure in the algorithm. In general, the integrated roll variable is replaced with an integrated pitch variable, which utilizes front and rear oppositely placed wheels for extensible member positions, that is, right front and right rear wheels instead of right front and left front wheels. Using these substitutions results in the pitch response desired compression and rebound values. An alternate diagonal comparison is also possible.

Block diagrams of the pitch response are shown in FIGS. 12a-12e which follow the flow of the roll response of FIGS, 11a-11e. Relative to FIGS, 11a-e, FIGS. 12a-e feature the various "R" subscripts replaced by "PH" subscripts to indicate the change from roll to pitch. Although the constants used in the pitch calculations can differ from those used in the roll calculations, in the particular implementation, the only changes to these are that Δ _PH = 1, and K_PHC = K_PHR = 1. Additionally, the equations are translated so that the various "LEFT" and "RIGHT" factors are replaced by "BACK" and "FRONT" factors. To correct for the sign of brake pressure, Block 1252 determines whether PH < 0 rather than the R≥ 0 determination of Block 1152. The use of a positive only brake pressure to indicate dive (with no indication of squat) precludes certain possibilities shown within FIGS. 12a-e. However, a full-flow diagram is shown to illustrate one possible use with a bi-directional longitudinal acceleration parameter.

Whereas the roll response is calculated in terms of separately calculating left and right responses for the front and rear of the vehicle, the pitch response is calculated by separately calculating front and back responses for the left and right sides of the vehicle. The other responses are typically computed serially for each of the four wheels.

U.S. Patent No. 4,634,142 discloses that the difference between integrated (low pass filtered) positions on opposite sides/ends are monitored to detect roll/pitch and opposing forces are applied to counteract body motions due to cornering or braking. It has been found that this works on reasonably flat roads, but bumpy roads caused poor response since the filter detection of roll and pitch becomes slower and less accurate. As disclosed above, forces/conditions are detected that would indicate that roll or pitch is about to occur before it does. Added optional sensors for this purpose by way of example these include roll and pitch accelerometers or alternatively a brake sensor, steering and speed sensors as disclosed in FIGS. 7a, e, f and 11a. This improved algorithm provides the correct counteracting forces before the motions even have a chance to occur. The integrated positions disclosed in the '142 patent now act as a feedback stabilizing loops to adjust the forces if and when motions try to occur as in FIGS, 11d, e. Additional algorithm improvements include: 1) modulation of the forces by vehicle speed FIG. 11b, 2) pre-filtering of the position to eliminate high frequency inputs to the DFT (Block 1132), and 3) small amplitude clipping and exponential filtering to try and help reduce harshness (rest of FIG. 11e). Simmlar comments apply to the pitch control of FIG. 12.

Bottoming and topping out response is provided to resist the tendency of a vehicle to bottom or top out. The need to provide forces to counter bottoming and topping out is related both to the displacement of the suspension from its equilibrium position and the velocity with which the suspension is travelling farther away from that position. The control system of the present invention provides for the addition of preventative compression and rebound forces to respectively resist bottoming and topping out once a suspension damper has travelled beyond a central zone or window approximately centered about the equilibrium position P_LTI. The damper extension corresponding to the upper or compression boundary of the central zone is identified as P_BOT (2 inches in the particular embodiment). Similarly, the damper extension corresponding to the lower or rebound boundary is P_TOP (-1-75 inches in the particular embodiment). As a further feature, however, once a preventative force has been applied so that the damper is no longer approaching a bottoming or topping out condition, an opposite corrective force, in rebound or compression, respectively, is applied to counter the impulse applied to the vehicle by the preventative force. As is shown in FIG. 13, at each cycle, the damper position P and damper velocity V_P are used to calculate preliminary forces BOT and TOP, the magnitude of which indicate the force appropriate to resist impending bottoming and topping out (Block 1310). These forces have components proportional to both the damper displacement relative to a respective window boundary and the damper velocity. In particular:

BOT = (P - P_BOT)K_PB/T + V_PK_VB/T

TOP = -[(P - P_TOP)K_PB/T + V_pK_VB/T] wherein the constants K_PB/T and K_VB/T are, respectively, damper position and velocity force constants which, in the described embodiment have values of 8 and 24. Optionally, differing constants can be provided for bottoming and topping out, respectively.

After the preliminary forces are calculated, the damper is determined to be in compression or rebound (Diamond 1312). If damper velocity V_P is greater than or equal to 0, the suspension is deemed under compression and, if less than 0, under rebound. If the suspension is in compression, it is next determined whether the damper is contracted beyond the upper boundary P_BOT of the window (Diamond 1314) by determining whether P is greater than P_BOT. if so, the suspension is viewed as about to bottom out and a compression force F_CB/T is provided which is equal to BOT and an impulse force F_I (the purpose of which is described in detail hereinbelow) is also set to BOT while a rebound force F_RB/T is set to zero. (Block 1316). If the damper is not contracted beyond the window, it is determined whether there is a need to counter an impulse arising out of applying a rebound force to resist previously topping out (Diamond 1318). This is achieved by determining whether an integrated impulse force (derived from F_I as is described hereinbelow) F_IINT is less than 0. If so, F_CB/T is set to the absolute value of F_IINT while F_RB/T and F_I are set to 0 (Block 1320). However, if F_IINT is not less than 0 and diere is thus no need to counter a prior impulse, F_CB/T, F_RB/T and F_I are all set to 0 (Block 1322).

Similarly, if the suspension is in rebound, it is next determined whether the damper is extended beyond the lower boundary P_TOP of the window (Diamond 1324) by determining whether P is less than P_TOP, the latter of which is typically a negative value. If so, the suspension is viewed as about to top out and a rebound force F_RB/T is provided which is equal to TOP, the impulse force F_I is set to negative TOP and the compression force F_CB/T is set to 0 (Diamond 1326). If me damper is not extended beyond the window it is determined whether there is a need to counter an impulse arising out of previously applying a compression force to resist bottoming out (Diamond 1328) which is done by determining whether F_IINT is greater than 0. If so, F_RB/T is set to F_IINT while F_CB/T and F_I are set to 0 (Block 1330). If F_IINT is not greater than 0 and diere is thus no need to counter a prior impulse, F_CB/T, F_RB/T an F_I are all set to 0 (Block 1340).

At the end of each such cycle, there results desired compression and rebound forces in the bottoming and topping out response, F_RB/T and F_CB/T. There is also a short term integration of the impulse force F_I to yield a new integrated impulse force FIINT. In particular, F_I is compared to F_INTn-1 (Diamond 1350). If greater, F_IINT is increased by a constant Δ_I (Block 1352) and if less, F_IINT is decreased by the constant Δ_I (Block 1354).

In the described embodiment, the constant Δ_I is equal to 5.

The aforementioned pitch and roll responses are provided to resist pitch and roll moments due primarily to respective longitudinal and lateral accelerations. Pitch and roll moments can also be introduced when encountering a bump. For example, if the left front wheel of the vehicle hits a bump, there is a tendency for the vehicle to have a combined squat and roll right movement. This may be approximated as a rotation about the diagonal axis between the right front and left rear wheels compressing the right rear suspension. The stored energy response compensates by having the right rear damper apply a compression- resisting force. The mode associated with such movement is identified as me Stored Energy mode. This mode is especially important where the bump or other influence causes displacement at a frequency near the sprung natural frequency of the vehicle, as higher frequency bumps do not cause so significant a rotation. For each wheel, compression and rebound stored energy responses F_CSE and F_RSE, respectively, are determined based upon the position and velocity of both that wheel (the wheel under control or WUC) and the wheel diagonally opposite (diagonal wheel or DW).

U.S. Patent No. 4,634,142 discloses that the bottoming and topping forces are a function of how close and how fast the suspension is to reaching its limit, a major improvement in the damping of the resulting motions that occur. More specifically, the ' 142 patent discusses that when a big bump is encountered and forces are applied to avoid hitting the suspension limit (bottom or top), the mass is set into motion and the response is improved if a portion of the energy injected into the system is removed. The present invention is an improvement to the prior method in that it measures how much energy is injected during a bump and then remove it (see Blocks 1350, 1352, 1345 and diamonds 1318, 1328 with their connected blocks).

First, as shown in FIG. 14a, every N_SE cycles (2 in the described embodiment), P is subject to a short-term integration (Block 1410) to yield an integrated value P_INTSE for the current WUC. The integration constant Δ_SE is 3 in the described embodiment. The details of the integration and the counter are not shown, but follow those described with regard to the previous modes. After me short-term integration, P_INTSE is clipped and subject to an exponential filter. In particular, if me absolute value of P_JNTSE ιs less tlιan a clipping value SE_CLIP (0.5 inch in the particular implementation) (Diamond 1412) an intermediate stored energy parameter K_SEEXP is set to 0 (Block 1414). If the absolute value of P_JNTSE is not less than SE_CLIP, the position is exponentially filtered. In particular, if | P_INTSE | is greater than the low amplitude filter constant SE_EXP (3 inches in the described embodiment) (Diamond 1416), K_SEEXP is set to 1 (Block 1418). Otherwise, K_SEEXP is set to P_INTSE divided by SE_EXP, quantity squared (Block 1420). Preliminary compression and rebound forces for the WUC F_WUCC and F_WUCR are then calculated by taking the product of K_SEEXP and P_INTSE and farther multiplying by a stored energy compression or rebound constant K_SEC or K_SER, respectively (Block 1422). In the described embodiment, K_SER ⁼

K_SEC = 0.7.

The process of FIG. 14a is repeated for the diagonal wheel (DW) so as to yield preliminary compression and rebound forces F_DWC and F_DWR, respectively. As shown in FIG. 14b, the preliminary compression rebound forces for the WUC and DW are selectively combined to yield the final compression and rebound forces in the stored energy mode, F_CSE and F_RSE, respectively. If the integrated position P_INTSE of the WUC and DW are wittiin a window defined by limits ±SE_LIM (1 inch in the particular implementation), no compression or rebound forces are provided. Thus, if P_INTSE of the WUC are not greater than SE_LIM (Diamond 1430) and is not less than -SE_LIM (Diamond 1431) and not greater dian SE_LIM of the DW (Diamond 1432) and not less than -SE_LIM (Diamond 1433), F_CSE and F_RSE are set to 0 (Block 1434). If, however, the WUC P_INTSE ιs less tnan SE_LIM

(Diamond 1431), it is men determined whether the DW P_INTSE is less th an SE_LIM

(Diamond 1435). If not, it is then determined whether the DW P_INTSE is greater than SE_LIM (Diamond 1436). If so, for the WUC, R_SE is set to 0 and F_CSE is set to the absolute value of F_DWC (Block 1437). If, however, for the diagonal wheel P_INTSE is less than

-SE_LIM, then, for the WUC, F_CSE is set to 0 and F_RSE is set to the absolute value of F_DWR minus the absolute value of F_WUCR (Block 1438). If the calculated F_RSE would be less dian (Diamond 1439), it is set to 0 (Block 1440). If P_INTSE is greater than SE_LIM for the wheel under control (Diamond 1430) and for the diagonal wheel (Diamond 1042), for the WUC, F_CSE is set to the absolute value of F_DWC minus the absolute value of F_WUCC and F_RSE is set to 0 (Block 1444), however, as with F_RSE, if the calculation F_CSE is negative (Diamond 1445), it is reset to 0 (Block 1446). If the DW P_INTSE is not greater than SE_LIM

(Diamond 1442) and is less than -SE_LIM (Block 1448), F_CSE is set to 0 and F_RSE is set to the absolute value of F_DWR (Block 1450).

U.S. Patent No. 4,634,142 discloses that the stored energy control applies counteracting forces to control the body motions. The detailed algorithms disclosed herein implement stored energy control. An improvement over the '142 patent is to add small amplitude clipping and exponential filtering to help reduce harshness (see FIG. 14a and Blocks 1430-1433, 1442, 1448).

The float response provides a small constant for damper force with respect to position for when the amplitude of motion at the sprung natural frequency or the amount of actual roll and pitch exceed certain minimum values. The constant or "Coulomb" force is dependent upon the vehicle velocity. In the described embodiment, it is applied in the rebound direction only as rebound damping forces are less perceptible than compression damping forces from the point of view of a vehicle passenger. Force is applied only to the extent that the other response modes are not providing such force.

As shown in FIG. 15, each cycle, to determine whether there is motion near the sprung natural frequency, the current and previous values of the integrated position P_SNF and P_SNFn-1 are received and compared (Diamond 1512) for determining whether the damper is under rebound or compression. If P_SNF is greater than or equal to P_SNFn-1 , damper is not in rebound and therefore it is meaningless to apply a rebound resisting force. Thus, P_RFL is set to 0 (Block 1514) (F_CFL is always 0). If, however, P_SNF is less than P_{SNFn-1 '} the suspension is in rebound, in which case the integrated sprung natural frequency parameter A_INTSNF is retrieved (not shown) and compared to a threshold float amplitude FL_SNFLIM which is the minimum sprung natural frequency movement for which Coulomb damping is applied (0.25 inches in the described embodiment) (Diamond 1518). If A_INTSNF is less than FL_SNFLIM' then no Coulomb force is necessary to damp sprung natural frequency motion and, in similar fashion, it is determined whether Coulomb force is required to damp roll motion by retrieving the integrated actual roll R_INT (not shown) and comparing it to a threshold float amplitude FL_RLIM (0.25 inches in the described embodiment) (Diamond 1522). If R_INT is less man FL_{RLI M,} then no Coulomb force is necessary to damp roll motion and it is similarly determined whether a Coulomb force is necessary to damp pitch motion by retrieving the integrated actual pitch PH_INT (not shown) and comparing that value to a threshold float amplitude FL_PLIM (0.25 inches in the described embodiment) (Diamond 1526). If PH_INT is less than FL_PLIM, there is no need for a Coulomb force and F_RFL is set to 0 (Block 1514). If, however, A_INTSNF is greater than or equal to FL_{SNFI M} or R_INT is greater than or equal to FL_RLIM or PH_INT is greater than or equal to FL_PLIM, then a necessary Coulomb force COUL is determined by retrieving the vehicle velocity dependent gain factor K_SR (not shown) and multiplying it by a constant force FL_COUL (200 lbs. in the described embodiment) (Block 1530). COUL is men compared to the sum of the rebound forces due to the sprung natural frequency, unsprung natural frequency, roll and pitch, pumping down, bottoming and topping out, and stored energy modes (Diamond 1532). If COUL is less dian this sum, there is no need for additional float damping and F_RFL is set to 0 (Block 1514). If COUL exceeds this sum, there is need for additional damping and F_RFL is set to the difference between COUL and the sum (Block 1534) so that when summed, the net rebound force F_R will equal COUL.

Although the described embodiment utilizes steering angle and vehicle speed to approximate roll-inducing lateral acceleration and brake pressure to approximate or indicate dive-inducing longitudinal deceleration, other methods and dieir associated hardware are available. For example, lateral (roll) accelerometer 60 and longitudinal (pitch) accelerometer 61 (such as the EBG-125 Series by Entran) can be used as an alternative or a supplement as shown in FIG. 18. A longitudinal accelerometer can be utilized within the pitch mode to provide anti-squat forces upon acceleration in addition to the anti-dive forces provided upon braking. Similarly, the longitudinal acceleration can be obtained by differentiating the output of the speed sensor. Similarly, as a supplement to brake pressure, engine vacuum pressure can be used to indicate acceleration.

Referring to FIGS. 19 and 20, block diagrams of roll control response calculations and pitch control response calculations, respectively, are shown for use with the alternate computer-controlled system of FIG. 18. The block diagram of FIG. 19 corresponds to the block diagram of FIG. 11a, with the exception of roll acceleration factor R_A in place of vehicle speed S and steering angle θ (Block 1912). Likewise, vehicle-dependent gain factors K_RAC and K_RAR are used (Block 1914).

The block diagram of FIG. 20 corresponds to the block diagram of FIG. 12a. The brake pressure ρ is replaced by pitch acceleration factor P_A (Block 2012). Likewise, the vehicle-dependent gain factors K_PHAC and K_PHAR are used (Block 2014).

A C-language computer program is attached as an Appendix appearing before the claims. This program contains an implementation of me principles of the invention and is incorporated herein by reference.

While an embodiment of me invention has been described and illustrated, odier constructions will be apparent to those skilled in the art. The control system can be used with a variety of damper constructions other than that specifically shown. The flow diagrams are illustrative of a process and do not limit that process to a particular implementation. One skilled in the art would be able to appropriately modify the implementation for a variety of purposes. A variety of reorderings or simplifications can be made to me particular illustrated steps without substantively altering the process. In particular, where conditions or choice determine certain relative values for the various properties and factors, the steps of the process and associated equations can become significantly simplified. Additionally, the above can be used, when appropriate, in non- vehicular applications. It is therefore understood diat the spirit of the invention can be practiced otherwise than specifically described.

Claims

WHAT IS CLAIMED IS:

1. A controller for controlling at least one damper, said at least one damper for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the displacement between the spring and unsprung masses, and a regulator responsive to independent compression and/or rebound control signals for adjusting, respectively, compression and/or rebound resisting forces of the at least one damper between the masses, the controller comprising:

a receiver for receiving the position signals;

a processor for processing the received position signals for determining at least one independent parameter to be representative of a distinct state of motion of the masses, the at least one independent parameter representing the amplitude of said motion occurring at one or more selective frequencies and another parameter representing the actual amplitude of said motion;

said processor further being adapted for determining, responsive to each said independent parameter, for said at least one damper, a desired independent response of the at least one damper in bom the compression and/or rebound directions, for each said distinct state of motion; and

a signal generator responsive to received compression and rebound control signals for control of each said regulator.

2. The system of claim 1 wherein, for each said at least one damper, a compression response is determined if the masses are displaced beyond a normal position, said compression response being proportional to the magnimde of the displacement beyond me normal position and wherein a rebound response is determined if the damper is retracted below said normal position, said rebound response being proportional to the magnimde of the retraction of the damper below the normal position.

3. The controller of claim 2 wherein the independent parameter representing the amplitude of said motion occurring at one or more selective frequencies is used as a gain factor to modify the value of the parameter representing the actual amplitude of said motion.

4. The controller of claim 1 wherein the processing means comprises a digital processor adapted to sample the received signals rapidly at a first rate and for providing control signals rapidly at a second rate.

5. The controller of claim 1 wherein the motion at selective frequencies is determined by applying a narrow bandwidth filter to the position signals by utilizing a filter.

6. The controller of claim 1 wherein one of the selective frequencies at which motion is determined is a natural frequency of the sprung mass.

7. The controller of claim 1 wherein the parameter being a function comprised of the amplitude of motion determined at one or more selective frequencies is limited to a maximum value.

8. The controller of claim 1 wherein, for each said one or more damper, the actual amplitude of motion is determined by forming an integration of the position signals received by the receiver.

9. The controller of claim 8 wherein the integration of the received signals is a linear integration.

10. The controller of claim 8 wherein the integration of the received signals is a non-linear integration.

11. The controller of claim 1 wherein the amplitude of motion occurring at one or more selective frequencies is determined by performing an integration of an output of said Discrete Fourier Transform.

12. The controller of claim 11 wherein the integration of the position signals is performed only on a falling edge.

13. The controller of claim 1 wherein the desired independent response for compression and/or rebound is reduced by the amount already provided by anodier control function.

14. The controller of claim 1 wherein the parameter being a function comprised of the amplitude of motion determined at one or more selective frequencies has a zero gain factor if the amplitude of motion is below a minimum level.

15. A controller of claim 1 wherein one of the selective frequencies at which motion is determined is a natural frequency of the unsprung mass.

16. The controller of claim 15 wherein the amplitude of motion occurring at said at least one selective frequency is determined by performing an integration of an output of a filter.

17. The controller of claim 16 wherein the integration is performed only on a falling edge of the signal.

18. The controller of claim 1 wherein the desired independent response for compression and/or rebound is reduced at an accelerated rate below a certain minimum threshold.

19. The controller of claim 1 wherein a rebound response is provided if the independent parameter representing the amplitude of motion occurring at one or more selective frequencies exceeds a minimum threshold.

20. The controller of claim 19 wherein the rebound response is a small constant force.

21. The controller of claim 19 wherein the rebound response is eliminated if a control function already provides a greater damping response to control the damper.

22. A controller for at least one damper, each said at least one damper comprising a body and an extensible member for damping motion between a sprung mass and an unsprung mass in compression and rebound directions, a sensor for generating position signals representative of the extension of the extensible member, and a regulator responsive to independent compression and/or rebound control signals for adjusting, respectively, compression and/or rebound resisting forces presented by the at least one damping device between the sprung and unsprung masses, the control system comprising:

receiver for receiving the position signals;

a processor for processing the received position signals for determining at least one or more independent parameters each representative of a distinct state of motion of the masses, at least one of the parameters being a function comprised of the motion determining at least one selective frequency and wherein me function includes decreasing the at least one parameter in a constant progression when the amplitude of the motion at said one or more selective frequencies decreases;

said processor being further adapted for determining, responsive to each independent parameter, for each said at least one damper, a desired independent response, in both compression and rebound directions, for each said distinct state of motion; and

means for processing a signal generator for generating, the desired independent response for each damping device for each damping device, responsive to the compression and rebound control signals for each regulator.

23. The controller of claim 22 wherein the processor comprises a digital processor adapted to sample the received signals rapidly a plurality of times per second and for providing control signals rapidly a plurality of times per second.

24. The controller of claim 22 wherein the motion at selective frequencies is determined through the use of a filter.

25. The controller of claim 22 wherein one of the selective frequencies at which motion is determined is a natural frequency of the sprung mass.

26. The controller of claim 22 wherein one of the selective frequencies at which motion is determined is a natural frequency of the unsprung mass.

27. A controller for controlling at least one damper, said at least one damper comprising a body and an extensible member for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the extension of the extensible member, and a regulator responsive to independent compression and/or rebound control signals for adjusting, respectively, compression and/or rebound resisting forces of the at least one damper between the masses, the controller comprising:

a receiver for receiving the position signals;

a processor for processing the received position signals for determining at least one independent parameter to be representative of a distinct state of motion of the masses, the at least one independent parameter representing an amplitude of motion determined by performing an integration of the position signals received by me receiver;

said processor further being adapted for determining, responsive to an actual amplitude of motion parameter, for each said at least one damper, a desired independent response of the damper in both the compression and rebound directions, for each distinct state of motion; and

a signal generator responsive to the received compression and rebound control signals for control of each said regulator.

28. The controller of claim 27 wherein, for each said at least one damper, a compression response is subtracted if the damper is extended, on average, beyond a normal position, said compression response being proportional to the extension beyond said normal position and wherein a rebound response is subtracted if the damper is retracted, on average, beyond said normal position, said rebound response being proportional to the retraction of the damper beyond said normal position.

29. The controller of claim 28 wherein the processor comprises a digital processor adapted to sample me received signals rapidly at a first rate and for providing control signals rapidly at a second rate.

30. A controller for controlling at least one damper, said at least one damper comprising a body and an extensible member for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the extension of the extensible member, at least one sensor for generating motion signals representative of horizontal motion of the vehicle and a regulator responsive to independent compression and/or rebound control signals for adjusting, respectively, compression and/or rebound resisting forces of the at least one damper between the masses, the controller comprising:

a receiver for receiving the position and motion signals;

a processor for processing the received position signals for determining at least one independent parameter to be representative of a distinct state of motion of the masses, the at least one independent parameter representing the magnimde and direction of change in the speed of the vehicle and the magnimde and direction of change in me attimde of the vehicle; said processor further being adapted for determining, responsive to each said independent parameter and to said motion signals, for said at least one damper, a desired independent response of the at least one damper in both the compression and rebound directions, for each said distinct state of motion; and

a signal generator responsive to received compression and rebound control signals for control of each said regulator.

31. The controller of claim 30 wherein said motion signals comprise a forward speed signal.

32. The controller of claim 30 wherein said motion signals comprise lateral and/or longitudinal acceleration signals.

33. The controller of claim 30 wherein compression and rebound responses are generated so as to counteract a tendency of the vehicle to change its attimde in response to changes in magnimde and direction of the vehicle speed.

34. The controller of claim 30 wherein me processor comprises a digital processor adapted for sampling the received motion signals rapidly at a first rate and for providing said control signals rapidly at a second rate.

35. The controller of claim 30 wherein the magnimde and direction of a change in attimde of the vehicle is determined by performing an integration of the received motion signals.

36. The controller of claim 30 wherein the magnimde and direction of change in speed of the vehicle is determined by performing an integration of the received motion signals.

37. The controller of claim 36 wherein the compression and rebound responses are decreased as the speed of the vehicle decreases.

38. The controller of claim 36 wherein the compression and rebound responses are eliminated as the speed of the vehicle decreases below a minimum speed.

39. The controller of claim 36 wherein the independent parameter representing the magnimde and direction of change in the attimde of the vehicle is reduced at an accelerated rate below a minimum threshold.

40. The controller of claim 36 wherein the independent parameter representing the magnimde and direction of change in the speed of the vehicle is reduced at an accelerated rate below a minimum threshold.

41. The controller of claim 40 wherein the independent parameter representing the degree and direction of change in the speed of the vehicle is eliminated below a second minimum threshold.

42. A controller for controlling at least one damper, said at least one damper comprising a body and an extensible member for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the extension of the extensible member, and a regulator responsive to independent compression and/or rebound control signals for adjusting, respectively, compression and/or rebound resisting forces of the at least one damper between the masses, the controller comprising:

a receiver for receiving the position signals;

a processor for processing the received position signals for determining at least one independent parameter to be representative of a distinct state of motion of the masses, the at least one independent parameter representing the velocity of said motion and another parameter representing the actual amplitude of said motion; said processor further being adapted for determining, responsive to each said independent parameter, for said at least one damper, a desired independent response of the at least one damper in both the compression and rebound directions, for each said distinct state of motion; and

a signal generator responsive to received compression and rebound control signals for control of each said regulator.

43. The controller of claim 42 wherein compression and rebound responses are generated so as to counteract a tendency of each damping device to reach its contraction and expansion travel limits, respectively.

44. The controller of claim 43 wherein an amount of energy imparted to the vehicle while preventing the tendency of me damper to reach its contraction and extension travel limits is measured and wherein compression and rebound responses are generated so as to remove said energy.

45. The controller of claim 43 wherein the compression and rebound responses so as to counteract the tendency of the damper to reach its contraction and extension travel limits is inhibited widiin an intermediate range of damper extension wherein no control is required.

46. The controller of claim 43 wherein the processor comprises a digital processor adapted for sampling the received position signals rapidly at a first rate and for providing controls signals rapidly at a second rate.

47. A method of controlling one or more damping devices, each damping device comprising a body and an extensible member for damping between a sprung mass and an unsprung mass in compression and rebound directions, a sensor for generating signals representative of the degree of the extension of the extensible members, and a regulator responsive to independent compression and rebound signals for adjusting, respectively, the compression and rebound forces presented to the shock absorber means to the extension, the method comprising:

receiving the position signals;

processing the received signals to determine one or more independent parameters for each of the distinct states of motion, at least one of me parameters being a function comprised of the motion determined at one or more selective frequencies and the motion determined wherein motion at lower frequencies is weighed greater dian motion at higher frequencies; determining, responsive to each independent parameter, for each damping device, a desired independent response for each distinct state of motion; and

generating compression and rebound control signals for each of the damping devices from the desired independent response for each of the regulators.

48. The method of claim 47 wherein the processing means comprises a digital processor adapted to sample the received signals at a first rate and for providing control signals rapidly at a second rate.

49. The memod of claim 47 wherein the motion at selective frequencies is determined by applying a narrow bandwidth filter to the position signals by utilizing a Discrete Fourier Transform.

50. The method of claim 47 wherein one of the selective frequencies at which motion is determined is a natural frequency of the sprung mass.

51. The method of claim 47 wherein me motion at lower frequencies is determined through the use of a low pass filter.

52. The method of claim 51 wherein the low pass filter is an integrator.

53. The method of claim 52 wherein the integrator is a non-linear integrator.

54. A controller system for a vehicle containing a plurality of damping devices, each damping device comprising a body and an extensible member for damping motion between sprung and unsprung masses, a sensor for generating a position signal representative of the degree of extension of the extensible member, and a regulator responsive to control signals for adjusting the compression and rebound resisting forces presented by the damping device, the control system comprising:

means for receiving the position signals;

means for receiving signals from the vehicle representative of vehicle velocity; means for processing the received signals to determine one or more independent parameters representative of the state of motion of the sprung and unsprung masses with respect to each other, widi at least one parameter being a function comprised of the velocity of the vehicle;

means for using each of the parameters to determine a desired independent response for each damping device; and

means for processing the desired independent responses for each of the damping devices for generating compression and rebound control signals for each of the regulator means.

55. The controller of claim 54 wherein the one parameter is a linear function of speed.

56. The controller of claim 54 wherein the one parameter is limited to a minimum value.

57. A controller for controlling one or more damping devices in a vehicle, each damping device comprising a body and an extensible member for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the degree of the extension of the extensible member, and a regulator responsive to independent compression and rebound control signals for adjusting, respectively, compression and rebound resisting forces presented by the device between the two masses, the control system comprising:

receiver means for receiving the position signals;

processor means for processing the position signals for determining one or more independent parameters, each independent parameter representative of a distinct state of motion of the two masses, at least one of the independent parameters being a function comprised of a signal representative of the velocity of the vehicle and a signal representative of the change in angular direction of the vehicle;

said processor means further determining, responsive to each independent parameter, for each damping device, a desired independent response in compression and rebound, for each distinct state of motion; and

generator means for generating the compression and rebound control signals for each regulator.

58. The controller of claim 57 wherein the signal representative of the change in angular direction of the vehicle is a function of the angle commanded by me steering mechanism of the vehicle.

59. A controller for controlling one or more damping devices, each damping device comprising a body and an extensible member for damping between sprung and unsprung masses in compression and rebound directions, a sensor for generating position signals representative of the degree of the extension of the extensible member, and a regulator responsive to independent compression and rebound control signals for adjusting, respectively, compression and rebound resisting forces presented by the device between the two masses, the control system comprising:

receiver means for receiving me position signals;

processor means for determining, for each damping device, a desired independent response in compression and rebound directions, said responses being proportional to the difference between the received position and the long term average of the received position; and

generator means for generating the compression and rebound control signals for each regulator.

60. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses in both compression and rebound directions, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, me controller comprising:

a processor responsive to signals representative of the position signals for forming at least one of me compression and rebound control signals for the regulator as a function of amplitude of motion between the masses and further as a function of amplitude of motion, for at least one selected frequency, of at least one of the masses.

61. The controller of claim 61 wherein the processor comprises means for forming at least one of compression and rebound signals as a function of a product of the amplitude of motion between the masses and me amplitude of motion, at at least one selected frequency, of at least one of me masses.

62. The controller of claim 60 wherein, for each said at least one damper, said processor comprises at least one of:

means for adjusting the compression control signals if the displacement of the masses is extended beyond a normal separation, said adjusted compression control signals being proportional to the magnimde of such extension beyond the normal separation; and means for adjusting the rebound control signals if the displacement of the masses is retracted beyond the normal separation, said adjusted rebound control signals being proportional to the magnimde of the retraction of the masses beyond such normal separation.

63. The controller of claim 60 wherein the masses have a relative velocity and the processor comprises means for forming such control signals also as a function of the amplitude of motion, for the at least one frequency, and the relative velocity between such masses.

64. The controller of claim 60 wherein the selected frequency is a natural frequency of such sprung mass.

65. The controller of claim 60 comprising a band pass filter and the processor passes signals representative of the amplitude of motion, for the at least one frequency, through the filter, in forming the control signals.

66. The controller of claim 65 wherein the filter is a Discrete Fourer Transform.

67. The controller of claim 65 wherein the processor comprises means for integrating the signal passed through the filter in determining the control signals.

68. The controller of claim 65 wherein the means for integrating integrates only on the falling edge the output from the filter.

69. The controller of claim 60 wherein the processor comprises means for limiting the control signals to a predetermined maximum.

70. The controller of claim 60 wherein the processor comprises means for limiting the plural means for adjusting the control signals and means for limiting the control signals by any of the plural means so that control signals do not exceed that adjusted by another of the plural means selectable predetermined maximum.

71. The controller of claim 60 wherein the at least one of the compression and rebound control signals is the rebound control signal.

72. The controller of claim 1 wherein me system also includes a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, and a sensor for generating motion signals representative of horizontal motion of the vehicle and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, the controller comprising:

the processor is responsive to signals representative of the motion signals for forming at least one of me compression and rebound control signals for the regulator also as a function of velocity of the vehicle.

73. The controller of claim 60 wherein the system includes a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, a sensor for generating motion signals representative of horizontal motion of me vehicle and the processor is responsive to signals representative of the position and motion signals for forming at least one of the compression and rebound control signals for the regulator as a function of the difference between the displacement between the masses at one time and the average of the displacement between the masses over a period of time.

74. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses in bom compression and rebound directions, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, the controller comprising:

a processor responsive to signals representative of the position signals for forming at least one of the compression and rebound control signals for the regulator as a function of amplitude of motion, at at least one selected frequency, of at least one of the masses said with the magnimde of such control signals decreasing in a constant progression as, amplitude of such motion decreases.

75. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses in both compression and rebound directions, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, the controller comprising:

a processor responsive to signals representative of the position signals for forming bom the compression and rebound control signals for the regulator as a function of an integration of the position signals.

76. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses of a vehicle in bom compression and rebound directions, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, a sensor for generating motion signals representative of motion of the vehicle and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, me controller comprising:

a processor responsive to signals representative of the position and motion signals for forming at least one of the compression and rebound control signals for the regulator as a function of amplitude of motion between the masses and further as a function of magnimde and direction of change in attimde of the vehicle.

77. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses of a vehicle in bom compression and rebound directions, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, a sensor for generating position signals representative of the displacement between the sprung and unsprung masses, and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, the controller comprising:

a processor responsive to signals representative of the position signals for forming at least one of the compression and rebound control signals for the regulator as a function of amplitude of motion between the masses and further as a function of velocity motion of the masses.

78. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses of a vehicle in bom compression and rebound directions, a sensor for generating position signals representative of the displacement between me sprung and unsprung masses, and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of me at least one damper between the masses, the controller comprising:

a processor responsive to signals representative of the position signals for forming at least one of the compression and rebound control signals for the regulator as a function of at least one selected frequency of motion of at least one of the masses.

79. A controller for controlling a damping system, the system having at least one damper for damping between sprung and unsprung masses of a vehicle in both compression and rebound directions, for a sensor for generating motion signals representative of horizontal motion of the vehicle and a regulator responsive to at least one of independent compression and rebound control signals for adjusting, respectively, at least one of compression and rebound resisting forces of the at least one damper between the masses, the controller comprising: a processor responsive to signals representative of the position and motion signals for forming at least one of the compression and rebound control signals for the regulator as a function of velocity and change in angular direction of movement of the vehicle.