A kind ofly lay the UUV method of adaptive fuzzy sliding mode control under disturbing by force in load
Technical field
The present invention relates to and lay the UUV method of adaptive fuzzy sliding mode control under disturbing by force in load.
Background technology
UUV good concealment, penetration ability are strong, if carry the load that can complete some particular task (torpedo of combat duty, submarine mine can be completed, small-sized UUV, video camera, the sonar of investigation tasks can be completed) and complete in specific region and lay, reach abrupt object surely, and the task that other modes can not realize can be realized.
The control laid under strong disturbance of UUV load belongs to non-linear, and very complicated, and sliding formwork controls the application being well suited for this process, but simple sliding formwork controls shake, and larger effect is not ideal, easily to UUV generation infringement physically.
Summary of the invention
The present invention does not lay studying a question of the control of the UUV under disturbing by force for load in the control method in order to solve current UUV, and a kind of of proposition lays the UUV method of adaptive fuzzy sliding mode control under disturbing by force in load.
A kind ofly lay the UUV method of adaptive fuzzy sliding mode control under disturbing by force in load and realize according to the following steps:
Step one: UUV carries out load and lays, and produces two kinds of interference to UUV, a kind of be load move in palisade pipe produce interference, a kind of be moisturizing cabin water inlet produce interference;
Step 2: obtain UUV current state μ, builds the kinetic model of UUV under load lays disturbance;
Step 3: according to step 2 design sliding-mode surface s, structure sliding mode controller;
Step 4: according to the sliding mode controller design fuzzy controller of step 3 structure, the input of fuzzy controller is sliding-mode surface s, and output is △ K, described △ K is the increment size of the switch control rule rule coefficient of sliding mode controller;
Step 5: utilize adaptive algorithm to optimize △ K, obtain
Step 6: step 5 is obtained
export to the sliding mode controller of step 3 structure, obtain new adaptive fuzzy sliding mode controller τ;
Step 7: the new adaptive fuzzy sliding mode controller τ control UUV utilizing step 6 to obtain, makes UUV state change;
Step 8: re-execute step 2 to step 7, until UUV reaches expectation state μ
dtill.
Invention effect:
The present invention adopts Adaptive Fuzzy Control to control handoff gain, there is the rapidity of external disturbance response, the adaptivity of external disturbance and inner parameter, and this controller significantly can reduce the buffeting of sliding mode controller, avoid because buffeting problem causes damage to UUV, make UUV complete after load lays and can return to appointment expectation state rapidly.
Accompanying drawing explanation
Fig. 1 is that moisturizing cabin is distributed in position, load section both sides schematic side view; In figure 1 is UUV, and 2 is load, and 3 is that moisturizing freight space is put;
Fig. 2 is the controller architecture figure based on Adaptive Fuzzy Sliding Mode Control;
Fig. 3 is UUV north orientation graph of errors analogous diagram;
Fig. 4 is UUV east orientation graph of errors analogous diagram;
Fig. 5 is UUV depth error curve analogous diagram;
Fig. 6 is UUV list error curve analogous diagram;
Fig. 7 is UUV trim error curve analogous diagram;
Fig. 8 is UUV yawing graph of errors analogous diagram.
Embodiment
Embodiment one: a kind ofly lay the UUV method of adaptive fuzzy sliding mode control under disturbing by force in load and comprise the following steps:
Step one: UUV carries out load and lays, and produces two kinds of interference to UUV, a kind of be load move in palisade pipe produce interference, a kind of be moisturizing cabin water inlet produce interference;
Step 2: obtain UUV current state μ, builds the kinetic model of UUV under load lays disturbance;
Step 3: according to step 2 design sliding-mode surface s, structure sliding mode controller;
Step 4: according to the sliding mode controller design fuzzy controller of step 3 structure, the input of fuzzy controller is sliding-mode surface s, and output is △ K, described △ K is the increment size of the switch control rule rule coefficient of sliding mode controller;
Step 5: utilize adaptive algorithm to optimize △ K, obtain
Step 6: step 5 is obtained
export to the sliding mode controller of step 3 structure, obtain new adaptive fuzzy sliding mode controller τ;
Step 7: the new adaptive fuzzy sliding mode controller τ control UUV utilizing step 6 to obtain, makes UUV state change;
Step 8: re-execute step 2 to step 7, until UUV reaches expectation state μ
dtill.
Embodiment two: present embodiment and embodiment one unlike: in described step one, UUV carries out load and lays, and produces two kinds of interference be specially UUV:
(1) load is moved and is produced interference in palisade pipe;
The equation of motion of load in palisade pipe is:
Wherein said m
tfor quality of loads, v
tfor the movement velocity of load in pipe, λ
11for the additional mass of load in pipe in direction of motion, F
tfor load airscrew thrust, R
xfluid resistance suffered by load, F
mfor the mechanical friction resistance between load and power valve;
The relational expression of run duration t and load stroke l is obtained by formula (1):
The Trimming Moment that load negative buoyancy force causes UUV is obtained by formula (2):
τ
x1=F
z1(l
1+l)(3)
Wherein F
z1for load negative buoyancy force, l
1for load lays the distance of its barycenter front and initial point;
UUV is longitudinally subject to load screw propeller:
F
x1=-F
T+R
x+F
m(4)
(2) interference of moisturizing cabin water inlet generation;
Be located at t
nthe quality that moment has entered water in moisturizing cabin is m
0, water intake velocity is v
n, at t
n+1(small) quality of the water that moment enters in moisturizing cabin is dm
0, before entering moisturizing cabin, former system of particles speed should be External airflow field speed v
0, after two merging, the speed of whole system of particles is v
n+1, then t
nthe kinetic energy of moment whole system of particles is:
T
n+1the kinetic energy of moment whole system of particles is:
At [t
n, t
n+1] in, the kinetic energy of whole system of particles is:
Wherein (micro-) merit of whole system of particles being done for bonding force of dW, dW is:
Wherein said S is the equivalent cross-sectional area in moisturizing cabin, and P is t
ngas pressure intensity in moment moisturizing cabin, P
2for t
n+1gas pressure intensity in moment moisturizing cabin,
for (micro-) merit that gas original in moisturizing cabin does whole system of particles,
concrete form be:
Wherein P
0for the pressure of gas in the front moisturizing cabin that intakes, it is all generally standard atmospheric pressure; P
nfor the pressure of gas in moisturizing cabin after intaking, x is equivalent water-depth in moisturizing cabin, x
nfor t
nthe equivalent water-depth in moment, L is that γ is the adiabatic exponent of gas at adiabatic compression stage moisturizing cabin total length, and γ value is 1.4;
Because moisturizing cabin water intake velocity is very fast, whole load lays process and will complete so the water inlet work of moisturizing cabin should complete in two seconds in 2s, and obtaining the water inlet of moisturizing cabin by formula (9) to the perturbed force of the vertical generation of UUV is:
Moisturizing process in moisturizing cabin to the Trimming Moment that UUV causes is:
Wherein said F
furepresent buoyancy when moisturizing cabin is sky, l
2for centre of buoyancy, moisturizing cabin is apart from the distance of the former heart;
Obtain load according to formula (3), (4), (10) and (11) to lay period and to be rivals in a contest the disturbance that UUV produces because of load cloth:
Wherein
F
x=F
x1+△
1(15)
F
z=F
z1+F
z2(16)
Wherein Λ is the transition matrix that moving coordinate system is changed to fixed coordinate system, △
1represent unknown disturbances power, △
2represent unknown disturbances moment, τ
p, τ
q, τ
rfor longitudinal, horizontal, vertical disturbance torque.
Embodiment three: present embodiment and embodiment one or two unlike: obtain UUV current state μ in described step 2, building the kinetic model of UUV under load lays disturbance is:
Obtaining UUV current state by a series of sensors of UUV self is
describe the position of UUV under earth coordinates and attitude vectors, wherein ξ, η, ζ are longitudinal, horizontal, the vertical coordinate under fixed coordinate system,
for pitch angle, roll angle, yaw angle;
UUV kinetic model is:
Wherein χ=[μ, v, w, p, q, r]
twherein u, v, w is respectively longitudinal, transverse direction under moving coordinate system and vertical velocity, p, q, be roll angle, pitch angle, yaw angle speed under r difference moving coordinate system, M representative system inertial matrix, C (μ) representative system coriolis force centrifugal force matrix, D (μ) represents fluid damping matrix, L (μ) represents other hydrodynamic forces, hydrodynamic moment suffered by UUV, G (μ) represents the restoring force and countermoment that are caused by gravity, buoyancy, and τ represents the propelling power and boost torque that UUV propulsion system provides, τ
drepresent external disturbance power and disturbing moment;
Because UUV mainly navigates by water certain depth under water when laying load, the impact of the sea wind be subject to, wave is less, so the impact of sea wind, wave is ignored, and definition τ
d=τ
dc+ τ
dm, therefore the kinetic model of UUV under load lays disturbance is:
Wherein τ
dmfor laying the disturbance that load causes, τ
dcfor the environmental interference that ocean current causes;
transformation matrix
A is the transition matrix that the angle under fixed coordinate system is changed under moving coordinate system;
M
*(μ)=J
-T(μ)MJ
-1(μ)(20)
G
*(μ)=J
-T(μ)G(μ)(24)。
Embodiment four: one of present embodiment and embodiment one to three unlike: design sliding-mode surface s in described step 3, structure sliding mode controller detailed process be:
The state error of UUV is:
e=μ
d-μ(25)
Wherein said μ
dfor the expectation state of UUV;
Sliding-mode surface is:
Wherein said matrix H is positive definite diagonal matrix;
Offset for reaching the object that load lays the disturbance of generation, design sliding mode controller is:
Wherein said K (t) is diagonal matrix, that is:
K(t)=diag(k
1,k
2,...,k
6),k
j=max(a
dj)+λ
j,j=1,2,…,6(28)
Wherein said a
djfor the vector acceleration M that load disturbance causes
*(μ)
-1τ
dmin a jth element, parameter lambda
j>0.Introduce parameter lambda
i>0 ensures the stability of controller.The object introducing handoff gain K (t) is in order to compensate for disturbances item τ
dm, τ
dc, to guarantee that sliding formwork existence condition one meets surely.
Embodiment five: one of present embodiment and embodiment one to four unlike: the detailed process designing fuzzy controller in described step 4 is:
Utilize the uncertain part △ K of the handoff gain of adaptive control and fuzzy control adjustment sliding mode controller, to reach the object reducing system chatter.Because the chattering phenomenon in Sliding mode variable structure system is mainly caused by the discontinuous switching of sliding mode controller, thus the effective way that impair system is buffeted be exactly guarantee compensating disturbance while, reduce to switch the gain of item.Because distracter τ
dm, τ
dcbecome when being and have uncertain factor inside, so the method with adaptive characteristic must be used to adjust handoff gain, while making system energy stable, effectively can weaken buffeting.
(1), the input of fuzzy controller is sliding-mode surface s, for variable s
i(i=1,2), define Q fuzzy set A
i m(m=1,2 ..., Q);
(2) fuzzy rule IFs is designed
iisA
i mtHEN △ k
iisB
i m, wherein, m=1,2 ..., Q, i=1,2,3,4,5,6, A
i mand B
i mfor monodrome fuzzy set; Specific as follows:
Arranging seven fuzzy set: NB in formula represents negative large, and during NM representative is negative, NS representative is negative little, and it is just little that ZE represents zero, PS representative, and PM represents center, and PB represents honest, △ k
jfor the increment of switch control rule rule gain;
From analyzing above, when | s
i| when being worth larger, | △ k
i| higher value should be got to ensure
it is larger negative value.When | s
i| when being worth less, | △ k
i| smaller value should be got to ensure
get negative value;
(3) membership function adopts Gaussian function:
Use monodrome fuzzy device and the average defuzzifier in center to complete the structure work of fuzzy system, the output valve of fuzzy system is:
Wherein
Free parameter vector,
Fuzzy basic functions,
represent the weight of i-th sliding-mode surface in m rule;
Obtain △ K=diag (△ k
1, △ k
2..., △ k
6).
Embodiment six: one of present embodiment and embodiment one to five unlike: utilize adaptive algorithm to optimize △ K in described step 5, obtain
detailed process be:
switch control rule rule optimum gain k
jdestimated value,
k
jderror,
the incremental representation of estimated gain is:
In formula,
the estimated value of free parameter vector,
adjustable monodrome controling parameters,
adaptive law be set as:
Wherein said
for the learning rate of adaptive system;
Embodiment seven: one of present embodiment and embodiment one to six unlike: the process that in described step 6, profit obtains new adaptive fuzzy sliding mode controller τ is:
Obtain
export to sliding mode controller, obtain new adaptive fuzzy sliding mode controller:
Wherein
Embodiment one:
Getting UUV is longitudinally north orientation, is laterally east orientation, and UUV length 5.5m, width 2m, height 1m, quality of loads is m
t=150kg, loaded length is 2m, and density of sea water is 1040kg/m
3, load buoyancy is 1324N, and speed when load goes out pipe is 3.1m/s, and going out the pipe time is 1.49s, load and from the mechanical friction resistance F navigated between power valve
m=44N, the fluid resistance of motion R of load
x=338N, after emulating, the state error e=μ obtained
dthe curve of-μ is as shown in Fig. 3-Fig. 8, as can be seen from the UUV state error curve of Fig. 3-Fig. 8, the error of parameters reduces to level off to 0 all gradually, and namely the virtual condition of UUV levels off to the expectation state of UUV very soon, and therefore controller of the present invention has good control effects.