CN103645736A - Non-linear H infinite inverse optimization output feedback controller-based ship course motion control method - Google Patents

Abstract

The invention discloses a non-linear H infinite inverse optimization output feedback controller-based ship course motion control method. A ship course motion and wave interference filter is designed; on the basis of combination of ship course information collected in real time by a shipbone compass, course motion information generated by the effect of a ship rudder and sea wave disturbance information are estimated; a ship course motion H infinite inverse optimization controller that is designed based on a ship course local optimization H infinite controller is used to obtain an optimum control rudder angle by combining the course motion information obtained by estimation, so that the ship can move along a preset course; a control effect monitor is designed to record the course angle; and the control rudder angle of the ship in real time and whether the ship can track a ship course reference model is determined; and if the tracking effect is bad, the controller is adjusted timely.

Description

Based on non-linear H ∞the ship course motion control method of inverse optimization output feedback controller

Technical field

What the present invention relates to is a kind of ship course motion controller structure.

Background technology

For security, maneuverability and the economy that ensures that boats and ships ride the sea, it is an important subject in ship control field that course is controlled.Ship motion is nonlinear in essence, and its dynamic perfromance depends on many factors such as ship type, ship's speed, the depth of water, load, drinking water, and these factors cause Ship dynamic situation model parameter to deviate from nominal value.Thereby ship course controller should design based on nonlinear control theory, and must there is the robustness to model parameter perturbation.Many advanced control algorithms, if H ∞ control, Based Intelligent Control, adaptive control, change structure control and contragradience control etc. are in succession for the design of direction controller, have been obtained and have been enriched achievement in research.

Marine course control system only can be measured boats and ships course angle, during application state feedback, must introduce the course angle rate information that state filter needs to obtain controller.In addition, ship's navigation Yu Haiyang, wave is a kind of inevitable environmental interference.The single order wave disturbance frequency of high frequency is positioned at outside marine system bandwidth or approaches with it but within actuator bandwidth.This just means if compensate with steering wheel the ship motion that high frequency wave agitation causes, will cause invalid steering frequently and cause the excessive wear of steering wheel and extra energy consumption.Therefore, filtering wave high frequency components in measuring-signal, and in being subject to the course measured value of wave Color Noise, estimate the boats and ships low frequency movement state (course angle, course angle speed etc.) that state feedback controller needs, realize wave filtering extremely important.Modern high performance ship control system, for improving estimation effect, often carrys out modeling wave disturbance and adopts expansion state method to carry out state estimation with forming filter.But forming filter is matching wave disturbance spectrum completely, and its parameter often can not accurately be obtained.When the state filter based on out of true forming filter and uncertain ship motion modelling is realized wave filtering like this, finally show as uncertain filtering error.

Ship course is controlled the optimum steering criterion that requires the comprehensive balance such as course precision, energy consumption and ship's speed loss, and linear quadratic optimal regulator (LQR) becomes the algorithm that is early applied to ship course control.But LQR designs based on locally linear model, and the effective uncertainty of transaction module and external disturbance.Non-linear H _∞system optimizing control becomes the optimum steering of acquisition and tackles these probabilistic a kind of direct selections, non-linear H _∞hamilton-Jacobi-Ai Siaike (HJI) equation corresponding to optimal control is difficult to solve conventionally, but application inverse optimization method can be avoided solving of HJI equation, by making the undetermined cost functional relevant to controller minimum, the solution of seeking HJI equation is converted into the Lyapunov function problem of structure closed-loop system.Meanwhile, ship motion nonlinear characteristic just obviously manifests when motion amplitude is larger, and this makes linear optimization controller when boats and ships slightly move, have premium properties, and the controller architecture based on the design of nonlinear optimization control theory must be possessed this Local Property.

Summary of the invention

The object of the present invention is to provide a kind of control performance better based on non-linear H _∞the ship course motion control method of inverse optimization output feedback controller.

For achieving the above object, the technical solution used in the present invention comprises the following steps:

Step 1: actuator receives command heading, starts to carry out ship course motion control;

Step 2: utilize boat-carrying compass to measure in real time naval vessel course angle, to boat-carrying computing machine, utilize ship course motion and wave interference filter to estimate the ship course angle ψ that rudder effect produces course angle communication _lestimated value

the ship course angular speed r that rudder effect produces _lestimated value the high frequency course angle ξ being produced by wave disturbance _hestimation

the high frequency course angle speed ψ that wave disturbance produces _hestimated value

and the estimated value of wave low-frequency excitation amount n

identification by low-and high-frequency course kinematic parameter is separated, from be subject to the measuring-signal of wave disturbance reconstruct low frequency movement signal, realize wave filtering;

Related ship course motion and wave interference filter expression formula are

$\begin{array}{c}\stackrel{\·}{\widehat{\mathrm{\ξ}=A_w\hat{\mathrm{\ξ}+k_1}}\widetilde{\mathrm{\ψ}}}\\ \stackrel{\·}{\hat{n}}=-T_n^{-1}\hat{n}+k_2\widetilde{\mathrm{\ψ}}\\ \stackrel{\·}{\widehat{{\mathrm{\ψ}}_L}}=\widehat{r_L}+k_3\widetilde{\mathrm{\ψ}}\\ \stackrel{\·}{\widehat{r_L}}=\mathrm{\θ\φ}\left(\widehat{r_L}\right)+\mathrm{bu}+\hat{n}+k_4\widetilde{\mathrm{\ψ}}\\ \widetilde{\mathrm{\ψ}}=\widehat{{\mathrm{\ψ}}_L}+C_w\widehat{\mathrm{\ξ}}\end{array}----\left(1\right)$

In formula, $\mathrm{\ξ}={\left(\begin{array}{cc}{\mathrm{\ξ}}_H& {\mathrm{\ψ}}_H\end{array}\right)}^T,$

$A_w=\left(\begin{array}{cc}0& 1\\ -{\mathrm{\ω}}_n^2& {-2\mathrm{\ζ}}_n{\mathrm{\ω}}_n\end{array}\right),$ ω _nfor encounter frequency, ζ _nrepresent additional damping, and have ζ _n∈ [0.01,0.1]; for filtering error, and have

$\mathrm{\&theta;}=\frac{{\mathrm{\&theta;}}_1\widehat{r_L}+{\mathrm{\&theta;}}_2{\hat{r}}_L^3}{\mathrm{\&phi;}\left(\widehat{r_L}\right)},$ Wherein ${\mathrm{\&theta;}}_1=-\frac{n_1}{T},{\mathrm{\&theta;}}_2=-\frac{n_3}{T},$ $\mathrm{\&phi;}\left({\hat{r}}_L\right)=\left(\begin{array}{c}{\hat{r}}_L\\ {\widehat{r_L}}^3\end{array}\right);$ N ₃, n ₁for constant, the boats and ships of course unstable are had to n ₁<0, course-stability n ₁>0; $b=\frac{K}{T},$ U=δ, T is tracing ability time index, and K is rudder gain coefficient, and δ is for controlling rudder angle; $C_w={\left(\begin{array}{c}0\\ 1\end{array}\right)}^T;$ K ₁∈ R ², k ₂, k ₃, k ₄∈ R is filter gain coefficient, and meets $k_1=-2({\mathrm{\&xi;}}_i-{\mathrm{\&xi;}}_n){\mathrm{\&omega;}}_c/{\mathrm{\&omega;}}_n,k_2=2({\mathrm{\&xi;}}_i-{\mathrm{\&xi;}}_n){\mathrm{\&omega;}}_n,k_3={\mathrm{\&omega;}}_c,$ ζ wherein _i> ζ _nfor notch filter damping, ω _c> ω _nfor cutoff frequency; T _nfor shift time constant, T _n>>1 and T _n ^-1<<k ₂k ₄< ω _n< ω _c;

Step 3: by what estimate in step 2 to obtain transfer to based on ship course motion local optimum H _∞controller and the ship course motion H that designs _∞inverse optimization controller, forms the instruction of optimum control rudder angle, and then acts on steering wheel, produces actual rudder angle; Control the course angle and control rudder angle of effect monitoring part real time record boats and ships;

Related ship course motion local optimum H _∞controller is

wherein $x=\left(\begin{array}{c}{x}_0\\ {\widehat{\mathrm{\&psi;}}}_L\\ {\hat{r}}_L\end{array},-{\mathrm{\&psi;}}_d\right),{\stackrel{\&CenterDot;}{x}}_0={\widehat{\mathrm{\&psi;}}}_L-{\mathrm{\&psi;}}_d,A=\left(\begin{array}{ccc}0& 1& 0\\ 0& 0& 1\\ 0& 0& {\mathrm{\&theta;}}_1\end{array}\right),B=\begin{array}{c}\left(\begin{array}{c}0\\ 0\\ b\end{array}\right)\end{array},u=\mathrm{\&delta;},D=\left(\begin{array}{cc}0& 0\\ 0& k_3\\ 1& k_4\end{array}\right),$ $\tilde{f}\left(x\right)=\left(\begin{array}{c}0\\ 0\\ \mathrm{\&theta;\&phi;}\left({\hat{r}}_L\right)-{\mathrm{\&theta;}}_1{\hat{r}}_L\end{array}\right);$ ψ _dfor ideal course as outlined angle;

Related ship course motion H _∞inverse optimization controller is

u _l=-R ^-1B ^TPx（3）

In formula, u _lcontroller for cost function optimum; R is weighting coefficient matrix, P=P ^t>0;

Related cost function expression formula is

PA+A ^TP+P(γ ^-2DD ^T-BR ^-1B ^T)P+Q=0

In formula, Q is weighting coefficient matrix, and γ is given Optimal Disturbance Rejection index, and has γ>=γ ^*, γ ^*for non-Optimal Disturbance Rejection index;

Step 4: repeating step two, step 3, until boats and ships move according to prearranged heading.

A kind of based on non-linear H _∞the ship course motion control method of inverse optimization output feedback controller, is further characterized in that, the control effect monitoring part in step 3 can judge whether boats and ships can follow the tracks of boats and ships prearranged heading, if tracking error is larger, adjusts in time controller;

The related SYSTEM ERROR MODEL of supervisory system is

$\begin{array}{c}\stackrel{\&CenterDot;}{\tilde{x}}=\tilde{A}\tilde{x}+\tilde{B}\widetilde{r_L}+\mathrm{Ew}\\ \tilde{z}=\tilde{C}\tilde{x}\\ \stackrel{\&CenterDot;}{\widetilde{r_L}=\mathrm{\&theta;\&phi;}\left(r_l\right)}-\mathrm{\&theta;\&phi;}\left(\widehat{r_L}\right)+\widetilde{z_e}\end{array}---\left(5\right)$

Wherein

be respectively ξ, n, ψ _l, r _l, ξ ₀evaluated error;

$\tilde{A}=\left(\begin{array}{cc}{A}_0-K_e{C}_0& 0\\ -k_2{C}_0& -T_n^{-1}\end{array}\right),A_0=\left(\begin{array}{cc}{A}_w& 0\\ 0& 0\end{array}\right),k_e=\left(\begin{array}{c}{k}_1\\ k_3\end{array}\right),\tilde{B}=\left(\begin{array}{c}{B}_0\\ 0\end{array}\right),B_0=\left(\begin{array}{c}0\\ 1\end{array}\right),E=\left(\begin{array}{cc}{E}_{w0}& 0\\ 0& E_n\end{array}\right),$

$E_{w0}={\left(E_{w,}0\right)}^T,E_w=\left(\begin{array}{c}0\\ k_w\end{array}\right),$ K _wfor strength of turbulence coefficient; $\tilde{C}={\left(\begin{array}{c}{k}_4{C}_0\\ -1\end{array}\right)}^T,C_0={\left(\begin{array}{c}{C}_w\\ 1\end{array}\right)}^T;w=\left(\begin{array}{c}{w}_H\\ w_n\end{array}\right),$ W _n, w _hfor standard zero-mean white Gaussian noise.

The relative prior art of the present invention has following advantage and effect:

Beneficial effect of the present invention is: by setting up ship course motion and wave interference filter, can accurately estimate the caused course of rudder action movable information, the course motion causing with wave disturbance, has effectively avoided invalid steering and has caused the excessive wear of steering wheel and extra energy consumption; By design ship course motion H ∞ inverse optimization controller, generation can make the control rudder angle instruction of cost function optimum, has improved control performance; By controlling effect monitoring, adjustment control parameter, has improved the adaptivity of the inventive method under different motion condition in time, guarantees that the control accuracy under different condition is stable.

Accompanying drawing explanation

Fig. 1 the present invention propose the workflow diagram of ship course motion control method.

Fig. 2 the present invention propose the block diagram of system of ship course motion control method.

Embodiment

Below in conjunction with accompanying drawing, the present invention is described in further detail.

As shown in Figure 2, a kind of based on non-linear H _∞the ship course motion control method of inverse optimization output feedback controller, its system forms wave filter, the non-linear H of the mathematical description, ship course motion and the wave interference that have comprised ship course motion and wave interference _∞inverse optimization output feedback controller, four parts of control effect monitoring.

When method is carried out, ship course motion and wave interfere information are estimated in same wave filter.The mathematical description that described ship course motion and wave disturb completes the mathematical model of ship course motion and describes, the low-frequency excitation force and moments such as second order wave disturbance, ocean current and crossing current wind and boats and ships not modeling dynamically for the mathematical description of course motion effects, wave three work of mathematical description to the single order high frequency components of ship course motion; Described Seawave Filter builds by state filter, can obtain the estimated value of boats and ships low frequency movement state, estimation and the low-frequency excitation of the high frequency motion state that produced by wave disturbance are estimated.By the separation to low-and high-frequency motion, from be subject to the measuring-signal of wave disturbance reconstruct low frequency movement signal, realized wave filtering; Non-linear H _∞inverse optimization output feedback controller completes ship course local optimum H _∞controller design and ship course H _∞the design of inverse optimization controller; Control effect monitoring and complete course angle and the real-time monitoring of controlling rudder angle under the different speed of a ship or plane.

One, the description that ship course motion and wave disturb, specifically comprises following content:

1) non-linear can being described as of oceangoing ship course motion

ψ wherein _lfor the ship course angle that rudder action produces, δ is for controlling rudder angle, and T is tracing ability time index, and K is rudder gain coefficient, H _n it is the nonlinear characteristic that three rank polynomial expressions are described n ₃, n ₁for constant, the boats and ships of course unstable are had to n ₁<0, course-stability n ₁>0.Make r _lfor course angle speed, have

formula (1) is expressed as state space form

$\left\{\begin{array}{c}{\stackrel{\&CenterDot;}{\mathrm{\&psi;}}}_L=r_L\\ \stackrel{\&CenterDot;}{r_L}={\mathrm{\&theta;}}_1{r}_L+{\mathrm{\&theta;}}_2{r}_L^3+\mathrm{bu}\end{array}\right.$

Wherein ${\mathrm{\&theta;}}_1=-\frac{n_1}{T},{\mathrm{\&theta;}}_2=-\frac{n_3}{T},b=\frac{K}{T},$ u=δ。

2) with single order Markov process, the low-frequency excitation force and moments such as second order wave disturbance, ocean current and crossing current wind and the boats and ships impact that modeling is not dynamically moved for course that ship course moves suffered described,

Wherein n is low-frequency excitation, w _nfor standard zero-mean white Gaussian noise, E _nrepresent white noise amplitude size, T _n>0 is shift time constant.For guaranteeing the gradual property of n, T _n>>1.

3) with second order forming filter, describe the single order high frequency components of wave to ship course motion, establish ψ _hfor producing high frequency course by wave disturbance, move.Amount of orientation

wave disturbance model is

$\begin{array}{c}\stackrel{\&CenterDot;}{\mathrm{\&xi;}}=A_w\mathrm{\&xi;}+E_w{W}_H\\ {\mathrm{\&psi;}}_H=C_w\mathrm{\&xi;}\end{array}---\left(4\right)$

Matrix wherein $A_w=\left(\begin{array}{cc}0& 1\\ -{\mathrm{\&omega;}}_n^2& -2{\mathrm{\&xi;}}_n{\mathrm{\&omega;}}_n\end{array}\right),C_w={\begin{array}{c}\left(\begin{array}{c}0\\ 1\end{array}\right)\end{array}}^T,E_w=\left(\begin{array}{c}0\\ k_w\end{array}\right),$ WH is standard zero-mean white Gaussian noise, ω _nfor encounter frequency, ζ _n>0 represents additional damping, general desirable ζ _n∈ [0.01,0.1], K _wfor strength of turbulence coefficient.

Therefore, boat-carrying compass or other navigational system can be measured course angle information and be

ψ=ψ _L+ψ _H=ψ _L+C _wξ(5)

Two, ship course motion and wave interference filter and filtering error system thereof, be specially

1) filter construction.The ship course kinematic system with wave disturbance forming for formula (1) (1)～formula (5) (5), establishes ξ, n, ψ _l, r _lestimated value be respectively wave filter is

$\begin{array}{c}\stackrel{\&CenterDot;}{\widehat{\mathrm{\&xi;}=A_w\hat{\mathrm{\&xi;}+k_1}}\widetilde{\mathrm{\&psi;}}}\\ \stackrel{\&CenterDot;}{\hat{n}}=-T_n^{-1}\hat{n}+k_2\widetilde{\mathrm{\&psi;}}\\ \stackrel{\&CenterDot;}{\widehat{{\mathrm{\&psi;}}_L}}=\widehat{r_L}+k_3\widetilde{\mathrm{\&psi;}}\\ \stackrel{\&CenterDot;}{\widehat{r_L}}=\mathrm{\&theta;\&phi;}\left(\widehat{r_L}\right)+\mathrm{bu}+\hat{n}+k_4\widetilde{\mathrm{\&psi;}}\\ \widetilde{\mathrm{\&psi;}}=\widehat{{\mathrm{\&psi;}}_L}+C_w\widehat{\mathrm{\&xi;}}\end{array}----\left(6\right)$

Wherein $\mathrm{\&theta;}=\frac{{\mathrm{\&theta;}}_1\widehat{r_L}+{\mathrm{\&theta;}}_2{\hat{r}}_L^3}{\mathrm{\&phi;}\left(\widehat{r_L}\right)},$

for filtering error, k ₁∈ R ², k ₂, k ₃, k ₄∈ R is filter gain coefficient undetermined.In formula (6), the 1st, 3,5 formulas can be write as

$\begin{array}{c}\stackrel{\&CenterDot;}{\widehat{{\mathrm{\&xi;}}_0}}=A_0\widehat{{\mathrm{\&xi;}}_0}+B_0\widehat{r_L}+k_e\widetilde{\mathrm{\&psi;}}\\ \widetilde{\mathrm{\&psi;}}=C_0\widehat{{\mathrm{\&xi;}}_0}\end{array}---\left(7\right)$

Wherein $K_e=\left(\begin{array}{c}{k}_1\\ k_3\end{array}\right),\widehat{{\mathrm{\&xi;}}_0}=\left(\begin{array}{c}\widehat{\mathrm{\&xi;}}\\ {\widehat{\mathrm{\&psi;}}}_L\end{array}\right),A_0=\left(\begin{array}{cc}{A}_w& 0\\ 0& 0\end{array}\right),B_0=\left(\begin{array}{c}0\\ 1\end{array}\right),E_{w0}=\left(\begin{array}{c}{E}_w\\ 0\end{array}\right),c_0={\left(\begin{array}{c}{c}_w\\ 1\end{array}\right)}^T.$

2) filtering error system architecture.Error system will be as control system monitoring model.Get ξ, n, ψ _l, r _l, ξ ₀evaluated error be respectively

filtering error system is

Wherein

definition

get

have

formula (8) (8) is converted into

Wherein

matrix $\tilde{A}=\left(\begin{array}{cc}{A}_0-k_e{c}_0& 0\\ -k_2{C}_0& -T_n^{-1}\end{array}\right),$ Matrix $\tilde{B}=\left(\begin{array}{c}{B}_0\\ 0\end{array}\right),$ Matrix $E=\left(\begin{array}{cc}{E}_{w0}& 0\\ 0& E_n\end{array}\right),$ Matrix $\tilde{C}=\begin{array}{}\end{array}{\left(\begin{array}{c}{k}_4{C}_0\\ -1\end{array}\right)}^T,$ Matrix $w=\left(\begin{array}{c}{w}_H\\ w_n\end{array}\right).$

3) filter gain coefficient k ₁, k ₂, k ₃, k ₄meet the following conditions

With

_t ^n-1<<k ₂k ₄< ω _n< ω _c(11) time, ζ wherein _i> ζ _nfor notch filter damping, ω _c> ω _nfor cutoff frequency.Filtering error systematic (8) is strict passive.

By state filter (6), just can obtain boats and ships low frequency movement state ψ _l, r _lestimated value

the high frequency motion state ξ being produced by wave disturbance _h, ψ _hestimation and low-frequency excitation is estimated

by the separation to low-and high-frequency motion, from be subject to the measuring-signal of wave disturbance reconstruct low frequency movement signal, thereby realized wave filtering.

Three, non-linear H _∞inverse optimization output feedback controller structure comprises ship course local optimum H _∞controller and ship course H _∞inverse optimization controller, is specially

1) state filter is realized wave filtering and state estimation, in order to realize the tracking in course, control, get the controller that state filter the 3rd, 4 formulas form and design a model, and while considering that boats and ships ride the sea, headway changes owing to being disturbed and handling resistance the uncertainty causing.Definition respectively

ψ _dfor ideal course as outlined angle.In addition, as eliminating the slow robustness part that becomes disturbance of normal value steady-state offset and reply, in controller design process, introduce integral action.Defining integration state so just have

$\left\{\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_0=x_1\\ {\stackrel{\&CenterDot;}{x}}_1=x_2+d_1\mathrm{\&omega;}\\ \stackrel{\&CenterDot;}{x_2=\mathrm{\&theta;\&phi;}\left(x_2\right)+\mathrm{bu}+d_2}\mathrm{\&omega;}\end{array}\right.---\left(12\right)$

Vector in formula $d_1={\left(\begin{array}{c}0\\ 0\\ k_3\end{array}\right)}^T,d_2={\left(\begin{array}{c}1\\ 1\\ k_4\end{array}\right)}^T,\mathrm{\&omega;}=\left(\begin{array}{c}\mathrm{\&Delta;\&theta;\&phi;}\left(x_2\right)\\ \mathrm{\&Delta;bu}\\ \widetilde{\mathrm{\&psi;}}\end{array}\right),\mathrm{\&phi;}\left(x_2\right)=\left(\begin{array}{c}{x}_2\\ x_2^3\end{array}\right),$ Δ θ and Δ b are uncertain.

System (12) local linear is turned to

Wherein

u=δ.Obviously,

controlled.

For local linearization system H _∞controller, has completed ship course local optimum H _∞controller design process.

2) for the local linearization system shown in formula (13), given Optimal Disturbance Rejection index γ>=γ ^*, γ wherein ^*for non-Optimal Disturbance Rejection index.Solving equation

Q wherein, R is weighting coefficient matrix, obtains P=P ^t>0, can be in the hope of making the controller of cost function optimum be

u ₁=-R ^-1BTPx(15)

This controller is ship course overall situation H _∞inverse optimization controller, can make the ability of system disturbance suppression the strongest.

Four, control effect monitoring based on wave filter error model, the course angle of energy real time record boats and ships and control rudder angle, judge whether boats and ships can follow the tracks of ship course reference model, if can finish by trace routine, if tracking effect is bad, adjust controller parameter, as shown in Figure 1.

Workflow diagram of the present invention as shown in Figure 1, ship's navigation, when starting to turn to, start the working procedure of controller, mathematical description design ship course motion and wave interference filter based on ship course motion and wave interference, complete the separation of low-and high-frequency motion, from be subject to the measuring-signal of wave disturbance reconstruct low frequency course angle and course angle speed motor message.According to the course of course angle and course angle speed motion composition, control model, adopt non-linear H _∞inverse optimization method has designed ship course movement output feedback controller.Control effect monitoring part supervisory control device output in real time data, judge whether course angle can follow the tracks of with reference to course and control rudder angle whether in executable scope, if control good results, complete the design of controller, boats and ships continue to turn to, if it is dissatisfied to control effect, readjust control parameter, until the control good results of output.

The present invention inputs to reference model by ordered rudder angle, and reference model is exported with reference to course, the non-linear H of ship course motion _∞inverse optimization output feedback controller receives with reference to course information and wave filter output information, through computing output order rudder angle, steering wheel-rudder is exported actual rudder angle to hull according to rudder angle instruction, boats and ships start to turn to, the steering angle of output records and inputs to wave filter via compass, so form closed-loop control system, as shown in Figure 2.

Claims (2)

1. one kind based on non-linear H _∞the ship course motion control method of inverse optimization output feedback controller, is characterized in that, comprises the steps:

Step 1: actuator receives command heading, starts to carry out ship course motion control;

Step 2: utilize boat-carrying compass to measure in real time naval vessel course angle, to boat-carrying computing machine, utilize ship course motion and wave interference filter to estimate the ship course angle ψ that rudder effect produces course angle communication _lestimated value

the ship course angular speed r that rudder effect produces _lestimated value

the high frequency course angle ξ being produced by wave disturbance _hestimation

the high frequency course angle speed ψ that wave disturbance produces _hestimated value and the estimated value of wave low-frequency excitation amount n

identification by low-and high-frequency course kinematic parameter is separated, from be subject to the measuring-signal of wave disturbance reconstruct low frequency movement signal, realize wave filtering;

Related ship course motion and wave interference filter expression formula are

In formula, ξ=(ξ _hψ _h) ^t,

ω _nfor encounter frequency, ζ _nrepresent additional damping, and have ζ _n∈ [0.01,0.1];

for filtering error, and have

wherein

n3, n1 are constant, and the boats and ships of course unstable are had to n1<0, course-stability n1>0; u=δ, T is tracing ability time index, and K is rudder gain coefficient, and δ is for controlling rudder angle;

k ₁∈ R ², k ₂, k ₃, k ₄∈ R is filter gain coefficient, and meets k1=-2 (ζ _i-ζ _n) ω _cω _n, k ₂=2 (ζ _i-ζ _n) ω _n, k ₃=ω _c, ζ wherein _i> ζ _nfor notch filter damping, ω _c> ω _nfor cutoff frequency; T _nfor shift time constant, T _n>>1 and T _n ^-1<<k ₂k ₄< ω _n< ω _c;

Step 3: by what estimate in step 2 to obtain

transfer to based on ship course motion local optimum H _∞controller and the ship course motion H that designs _∞inverse optimization controller, forms the instruction of optimum control rudder angle, and then acts on steering wheel, produces actual rudder angle; Control the course angle and control rudder angle of effect monitoring part real time record boats and ships;

Related ship course motion local optimum H _∞controller is

Wherein

ψ _dfor ideal course as outlined angle;

Related ship course motion H _∞inverse optimization controller is

U _l=-R ^-1b ^tpx(3) in formula, u _lcontroller for cost function optimum; R is weighting coefficient matrix, P=P ^t>0;

Related cost function expression formula is

PA+A ^tp+P (γ ^-2dD ^t-BR ^-1b ^t) P+Q=0(4) in formula, Q is weighting coefficient matrix, γ is given Optimal Disturbance Rejection index, and has γ>=γ ^*, γ ^*for non-Optimal Disturbance Rejection index;

Step 4: repeating step two, step 3, until boats and ships move according to prearranged heading.

2. as claimed in claim 1 a kind of based on non-linear H _∞the ship course motion control method of inverse optimization output feedback controller, is characterized in that, the control effect monitoring part in step 3 can judge whether boats and ships can follow the tracks of boats and ships prearranged heading, if tracking error is larger, adjusts in time controller;

The related SYSTEM ERROR MODEL of supervisory system is

Wherein

be respectively ξ, n, ψ _l, r _l, ξ ₀evaluated error;

k _wfor strength of turbulence coefficient;

w _n, w _hfor standard zero-mean white Gaussian noise.

Images