KR20160081906A - Filtering back-stepping ship motion control system based on adaptive fuzzy estimator - Google Patents

Abstract

The present invention seeks to provide a filtering back stepping ship motion control system based on an adaptive fuzzy estimator. The system comprises a control system 2, a guide system 4, a differential strain transducer 6, a data processing system 7 and a sensor system 12, The sensor system 12 includes a position and orientation sensor 11 and a velocity sensor 10 and the data processing system 7 acquires the expected position of the data fusion system 9, And a filtering system 8 and the control system 2 comprises a filtering back stepping controller 3 and an adaptive fuzzy estimator 5. The adaptive fuzzy estimator 5 comprises a guide system 4 and a non- The filtering back stepping controller 3 receives the data of the converter 6 at the same time, and the filtering back stepping controller 3 receives the expectation information provided by the guide system 4 and its derivative, the new state variable information provided by the differential transformer 6, The unknown linear function provided by the estimator 5 It receives the estimated output at the same time.

Description

TECHNICAL FIELD [0001] The present invention relates to a filtering back-stepping motion control system based on an adaptive fuzzy estimator,

The present invention relates to a ship motion control system.

In recent years, abundant marine resources have been attracting attention from each country, and more and more countries are increasingly interested in marine processing and marine development. As the maritime activities of people become more and more frequent, the technical field related to ship control has received a wide interest of researchers. Also, ship navigation direction control, navigation track control, route tracking and dynamic positioning control technology are also studied in the control field. It is becoming an issue.

Although many nonlinear control methods are based on model information in the design process of ship motion controller, it is very difficult to construct an accurate system model in general. Therefore, ship motion in the case where unknown information exists in the model It is very necessary to base the adaptive fuzzy estimator based nonlinear control method. In addition, the derivative of the virtual control amount should be obtained at every step in the process of designing the controller by the conventional backstepping method. If the order of the system increases, the complexity of obtaining the derivative is complicated, There is also a constant demand for the characteristics of the system. To simplify the design procedure of the controller by introducing a second-order filter to approximate the virtual control amount and its derivative to omit the step of obtaining the derivative of the virtual control amount in the process of designing the back stepping controller; The introduction of the filter can improve the suppression ability of the controller against the noise. Most of the current literature reports that dynamic control of a full driver ship, such as dynamic position control, does not take into account the dynamic characteristics of the propeller, but simply acts on the ship directly with control and torque, There is a certain difference.

Based on the results of literature review, Zhou Hongbo et al. In "Control and Strategy" (4th FY 2012), "Trajectory tracking control of unmanned helicopter based on filtering back stepping method" We designed a trajectory tracking controller according to the filtering back stepping method for the unmanned helicopter and simplified the design of the controller by approximating the virtual control amount and its derivative using a filter instead of directly deriving the virtual control amount. He Yuebang et al. (2009) presented a flight control design based on the "unmanned helicopter robust integral filtering back-stepping method" The application of the stepping method to the tracking of the unmanned helicopter trajectory has been extensively studied, and the anti-jamming ability of the closed loop system has been improved by introducing the integral term and the robust term. However, all of the above studies are related to systems already known to the model.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a filtering back stepping ship motion control system based on an adaptive fuzzy estimator which is related to the case where an unknown nonlinear function exists in a model.

An adaptive fuzzy estimator-based filtering back stepping system including a control system 2, a guidance system 4, a differential dynamics transformer 6, a data processing system 7 and a sensor system 12 according to the present invention, In the ship movement control system, the guide system (4) generates one smoothed path in accordance with the input expected value and the initial position of the ship, and calculates an expected position at each time of the ship And the sensor system 12 includes a position and orientation sensor 11 and a velocity sensor 10, wherein the actual position and orientation of the vessel collected from the position and orientation sensor 11, The direction of the bow is transmitted to the data processing system 7 together with the information of the speed of movement of the ship collected by the speed sensor 10 and the processing of the data fusion system 9 and the filtering system 8 in the data processing system 7 Through The low frequency position and attitude information and the speed information of the obtained ship are transferred to the differential decimating transducer 6, and a new state variable is obtained through the state change; The adaptive fuzzy estimator 5 transmits the new state variable to the filtering back stepping controller 3 and the adaptive fuzzy estimator 5 of the control system 2. The adaptive fuzzy estimator 5 receives the data of the guide system 4 and the differential- Estimates of the unknown nonlinear function required by the controller include estimates of low frequency interference as well; The filtering back stepping controller 3 calculates the difference between the expected information provided by the guide system 4 and its derivatives, the new state variable information provided by the differential transformer 6 and the unknown nonlinear function provided by the adaptive fuzzy estimator 5, And obtains the corresponding control command information through calculation to adjust the longitudinal thrust, the lateral thrust and the heading torque of the ship, thereby realizing accurate control of the ship.

The adaptive fuzzy estimator 5 simultaneously receives the data of the guide system 4 and the differential transformer 6 and the estimation of the unknown nonlinear function required by the controller is performed by an adaptive fuzzy estimator 5 based on the new state variables provided by the differential transformer 6 and the expected information provided by the guide system 4, the fuzzy logic system can be applied to the unknown nonlinear function and low- By performing the integrated estimation, it means acquiring the nonlinear function required by the filtering back stepping controller.

Further, a second-order filter is introduced into the filtering back stepping controller 3 in the control system 2, and approximates to the virtual control amount and its derivative through the filter.

The differential differential transducer 6 converts the position and attitude information and the speed information of the ship to obtain new state variable information and transmits the information to the control system 2.

In the control system 2, the filtering back stepping controller 3 combines the new state variable information according to the expected position and attitude provided in the guide system 4 and its derivative information, and receives the estimated value of the unknown nonlinear function And generates a control force and a control torque command through a series of nonlinear calculations, and controls the position and the navigation direction of the ship by driving the propulsion system.

Further, a second-order filter is introduced into the filtering back stepping controller 3 in the control system 2, and approximates to the virtual control amount and its derivative through the filter.

The main features of the present invention are implemented as follows.

1) The guide system 4 generates one smooth expected path by the given expected target and the initial position of the ship, acquires the expected position and attitude of the ship at each time based on the path, and its derivatives, Can be set at the expected position in the direction of the final expected through the equal speed, constant speed, etc. deceleration process; By transferring the obtained expected position and posture and its derivative to the control system 2, the unknown nonlinear function is estimated and the control command is calculated.

2) The sensor system 12 is divided into a position and orientation sensor 11 and a velocity sensor 10. These sensors transmit the collected information to the data processing system 7, .

3) The data processing system 7 includes a data fusion system 9 and a filtering system 8, and the data fusion system 9 acquires corresponding information by fusing data collected from a plurality of sensors And then filtered through the filtering system 8 to finally obtain the ship motion information to be used in the control system 2 and to transmit the obtained information to the differential dynamic transducer 6.

4) The differential differential transducer 6 obtains a new state variable through the differential state transition of the ship with respect to position and attitude information and velocity information provided by the data processing system 7, .

5) The control system 2 comprises two parts: an adaptive fuzzy estimator 5 and a filtering back stepping controller 3, and the adaptive fuzzy estimator 5 calculates the expected information provided by the guidance system 4, Estimates an unknown nonlinear function through a fuzzy system corresponding to a certain adaptation rule, and transmits the estimated nonlinear function to the filtering back stepping controller 3 to obtain a control command Calculate; The filtering back stepping controller 3 calculates an estimated value of the unknown linear function provided by the adaptive fuzzy estimator 5 and a new state variable provided by the differential deconvolution transformer 6 based on the expected information provided by the guide system 4 To obtain a suitable control command through calculation of a series of nonlinear control algorithms, to drive the execution device of the ship, and to adjust the position and speed of the ship to achieve the control objective.

The filtering back stepping controller 3 of the control system 2 introduces the second filter technique in the controller design process without approximating the derivative of the virtual control amount and approximates the virtual control amount and its derivative thereof through the filter, In addition to simplification, the noise suppression ability of the controller also improved.

An advantage of the present invention is that accurate control of ship motion can be realized without needing to know a mathematical model of accurate ship motion. The adaptive fuzzy estimator according to the design of the present invention not only estimates the unknown nonlinear function in the model but also includes an estimation procedure for external low frequency interference, thereby improving robust performance against interference of the system; The filtering back stepping method simplifies the form of the controller and simplifies the designing process by replacing the process of obtaining the derivative of the virtual control amount in a typical back stepping method by introducing a filter to approximate the virtual control amount and its derivative. The introduction of the filter also improved the noise suppression ability of the controller.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is an overall block diagram of a filtering back stepping ship motion control system based on an adaptive fuzzy estimator.
2 is a diagram showing a configuration of a filtering back stepping controller.
3 is a diagram showing a membership function curve of the adaptive fuzzy system.
4 is a view showing a position and a heading tracking error curve.
5 is a diagram showing a control command curve of the controller.

Hereinafter, the present invention will be described in detail.

FIG. 1 shows the overall configuration of a filtering back stepping ship movement control system based on an adaptive fuzzy estimator according to the present invention. In FIG. 1, the symbols have the following meanings. 1 - environmental interference; 2 - control system; 3 - filtering back stepping controller; 4 - Guide system; 5 - adaptive fuzzy estimator; 6 - differential diffeomorphism transducer; 7 - Data Processing System; 8 - Filtering system; 9 - Data fusion system; 10 - speed sensor; 11 - position and position sensor; 12 - Sensor system; 13 - represents the night.

1, a filtering back stepping vessel motion control system based on an adaptive fuzzy estimator according to the present invention includes a control system 2, a guidance system 4, a differential strain transducer 6, a data processing system 7, Sensor system 12. The position and orientation sensor 11 in the sensor system 12 collects the actual position and attitude information of the ship and packs it together with the movement information of the ship collected by the velocity sensor 10 and transmits the information to the data processing system 7 This data is processed by the data fusion system 9 and the filtering system 8 to obtain position and attitude information and speed information of the ship applied to the control system 2; The data after the processing is transferred to the differential decalculation converter 6 and acquires a new state variable to proceed the back stepping design through the state change; Transferring the new variable to the filtering back stepping controller 3 and the adaptive fuzzy estimator 5 of the control system 2 so as to proceed with the corresponding calculation in the control system; The adaptive fuzzy estimator 5 simultaneously receives the data of the guide system 4 and the differential transformer 6, and the estimation of the unknown nonlinear function required by the controller also includes an estimation of the low frequency interference; The filtering back stepping controller compares the expected information provided by the guide system 4 and its derivatives, the new state variable information provided by the differential decimating transducer 6, and the estimated output of the unknown nonlinear function provided by the adaptive fuzzy estimator 5 And obtains corresponding control command information through a series of calculations to achieve accurate control of the vessel by adjusting the longitudinal propulsion, transverse propulsion and heading torques of the vessel.

및

) 를 획득할 수 있다. 편이를 위하여,

,

로 표기할 수 있다.1) The guide system 4 automatically generates one smoothed path according to the set expected value ? _D and the initial position of the ship, and based on the path, (X _d , y _d ), the expected bow direction (ψ _d ), and its derivative (ψ _d ) at each time of the ship so as to reach the designated position

And

Can be obtained. For convenience,

,

.

)와 속도 정보(

)를 획득하고; 처리를 거친 데이터는 미분동형사상 변환기(6)로 전달되어 상태 변화되며, 좌표적 변환(

)을 통해 새로운 변수(x ₁, x ₂)를 획득하고, 이들로 기존의 시스템과 동등한 새로운 모형을 구성한다. 새로운 모형의 수요를 만족시키기 위하여, 새로운 기대 위치 및 자세와 속도를

,

로 표기한다.2) The sensor system 12 includes a position and orientation sensor 11 and a velocity sensor 10, and collects the actual position of the ship and the actual heading angle and velocity information, respectively. The position and attitude information and the speed information of the ship are transmitted to the data processing system 7 and the position of the ship which can be applied to the control system 2 after the processing of the data fusion system 9 and the filtering system 8 And attitude information

) And speed information (

); The processed data is transferred to the differential transformer 6 and changed in state, and the coordinate transformation

( X ₁ , x ₂ ), and construct a new model equivalent to the existing system. In order to satisfy the demand of the new model, new expectation position and attitude and speed

,

.

The model of the existing system is as follows:

와

는 실행 장치와 관련되는 계수 행렬을 나타낸다.In the above relation, η denotes the ship's position and bow direction vector, ν denotes the ship's velocity vector, b denotes low frequency interference capability, and J (η) denotes a transformation matrix between the ship's body coordinate system and the geodesic coordinate system. D denotes a damping matrix, τ denotes a control vector, τ _e denotes a control command vector, M denotes a system inertia matrix, C (ν) denotes a coriolis centrifugal force matrix, D (ν) denotes a damping matrix,

Wow

Represents the coefficient matrix associated with the execution unit.

The equivalent system obtained after differential differential crimp change is as follows:

Among the above relational expressions,

3) Design the following filtering back stepping controller based on the equivalent model obtained above:

은 2차 필터의 출력을 나타내고, 이는 각 가상 제어량에 근사하기 위한 것이며, 그 도함수인

도 2차 필터에 의해 출력된다.among them,

Represents the output of the second-order filter, which is to approximate each virtual control amount, and its derivative

Is also output by the secondary filter.

Here, the expected value of each virtual control amount is as follows:

이며; k _i(i = 1, 2, 3)는 제어 이득 행렬(양한정 대각 행렬)을 나타내고; ν _i(i = 1, 2, 3)는 각 추적 오차에 따른 보상 벡터를 나타내고, 아래와 같이 정의된다.among them,

; k _i (i = 1, 2, 3) represents the control gain matrix (positive definite diagonal matrix); ν _i (i = 1, 2, 3) represents a compensation vector according to each tracking error, and is defined as follows.

는 아래와 같이 정의된다. Among them, vector

Is defined as follows.

의 초기값은 0(

，i = 1, 2)이며,

이다. 이때 제어 법칙은, G ₁ = 1, g ₂ = M _? ^-1 , g ₂ = B , and

The initial value of 0 (

, I = 1, 2)

to be. At this time,

τ _e = α ₃ ,

와

(i = 1, 2, 3)는 아래와 같이 정의된다.Controller design process

Wow

(i = 1, 2, 3) is defined as follows.

,

;1) When i = 1,

,

;

와

는 필터에서 출력된다.2) When i = 2, 3,

Wow

Is output from the filter.

은 설정된 위치 확인 타겟

를 의미하고,

는 설정된 추적 속도

를 의미한다.Please note:

Is set to a configured location target

Lt; / RTI >

Lt; / RTI >

.

Each filter can be defined as follows.

와

는 유계이며 연속된다.In the above relation, I represents a cubic unit matrix. As can be seen, when ? _{(I-1) c} is a constant,

Wow

And is continuous.

4) All of the above-mentioned designs proceed under the assumption that the parameters of the model are already known, but it is very difficult to design the controller based on the model, in general, when the parameters of the model are unknown or some parameters are unknown. The adaptive fuzzy system is introduced to solve the unknown parameter problem of the model, and the unknown nonlinear function required for the controller is estimated.

Assuming that there are N rules in the fuzzy rule base, the i-th has the following form.

R ⁱ : IF x ₁ is μ ₁ ⁱ and ... ^{_{and x n is μ n i,}} then y is B i (i = 1, 2, ..., N)

Among them, μ _n ⁱ is a membership function of x _n (n = 1, 2, ..., N).

If so, the output of the fuzzy system can be expressed as:

,

를 퍼지 추정 파라미터 벡터로 정의하고,In the above relationship

,

Is defined as a fuzzy estimation parameter vector,

인 관계를 만족한다.

Lt; / RTI >

In order to approximate the unknown nonlinear function (f) required by the controller, the adaptive fuzzy system can be used to approximate the median elements of f sequentially. In other words,

Among them, the approximate function (?) Of the nonlinear function (f) can be defined as follows.

,

이다.among them,

,

to be.

The optimal estimation vector is defined as θ ^* and also satisfies the following condition for a given small positive value (ε (ε> 0)).

이다.among them,

to be.

일 경우, 적응률은 아래와 같이 선택된다.

, The adaptation rate is selected as follows.

(i = 1, 2, 3)

(i = 1, 2, 3)

Where r _i > 0, k _i > 0 are design parameters and ν _2i is the i-th element of ν ₂ .

,

일 경우, 이를 아래와 같이 정의한다.

,

, It is defined as follows.

,

,

Among them, I _n is an n-order unitary matrix. If so, the adaptation rate can be expressed in vector form as follows.

Therefore, the control system 2 can calculate the control command based on the formula τ _e = α ₃ , and can control the position and the heading direction of the ship.

In the present invention, simulation experiments are carried out by applying a nonlinear mathematical model of a surface ship, and the parameters of the ship model according to the simulation experiment are as follows.

among them,

,

,

,

,

이다.

to be.

In the simulation process, we add the following interference and uncertain parameters,

Among them, the parameter marked with "* " is a nominal model parameter.

The initial position coordinates of the parameter are (0m, 0m, 0deg), the initial velocity is (0m / s, 0m / s, 0deg / s) and the expected position is (2m, 1m, 5deg). The simulation results refer to Figs. 3 to 5.

Under the function of the motion controller of the filtering backstepping ship based on the adaptive fuzzy estimator proposed in the present invention through the simulation curve and data analysis, the ship can overcome the influence of the uncertain characteristics of the model parameter and the interference from the outside It can be seen that the expected position provided by the guide system can be quickly tracked to maintain a given bow direction and achieve the desired control effect under the action of a relatively smooth control force. This means that the designed adaptive fuzzy estimator can make a relatively good estimation of the unknown nonlinear function and interference of the ship model and the filter in the filtering back stepping method can be very close to the virtual control amount and its derivative, The process of obtaining the derivative of the virtual control amount is omitted, thereby simplifying the design process of the controller. As can be seen from the simulation results, the control law according to the present invention has global asymptotic tracking characteristics and has relatively good robustness against model parameter uncertainty and unmodeled dynamics.

Claims (3)

1. An adaptive fuzzy estimator based filtering back stepping ship motion control system comprising a control system (2), a guidance system (4), a differential strain transducer (6), a data processing system (7) and a sensor system (12)
The guide system 4 generates one smooth path based on the input expected value and the initial position of the ship and acquires the expected position, expected bow direction and expected speed at each time of the ship in accordance with the route; The sensor system 12 includes a position and orientation sensor 11 and a velocity sensor 10 that collects in the velocity sensor 10 the actual position and heading angle of the ship, Frequency position and attitude information of the ship obtained through the processing of the data fusion system 9 and the filtering system 8 in the data processing system 7 together with the moving speed information of the ship, And velocity information to differential differential transducer 6 and acquires a new state variable via state transitions; The adaptive fuzzy estimator 5 transmits the new state variable to the filtering back stepping controller 3 and the adaptive fuzzy estimator 5 of the control system 2. The adaptive fuzzy estimator 5 receives the data of the guide system 4 and the differential- Estimates of the unknown nonlinear function required by the controller include estimates of low frequency interference as well; The filtering back stepping controller 3 calculates the predicted position, the expected head direction, the expected speed provided by the guide system 4, the new state variable information provided by the differential decelerating transformer 6, By receiving the estimated output for the unknown nonlinear function at the same time, obtaining the corresponding control command information through calculation, and by adjusting the longitudinal thrust, the lateral thrust and the heading torque of the ship, Adaptive Fuzzy Estimator Based Filtering Backstepping Ship Motion Control System. The method according to claim 1,
The adaptive fuzzy estimator 5 simultaneously receives the data of the guidance system 4 and the differential decelerating transducer 6 and the estimation of the unknown nonlinear function required by the controller is performed by the adaptive fuzzy estimator 5 in the control system 2, Based on the new state variables provided by the differential transformer 6 and the expected information provided by the guidance system 4, the integrated nonlinear function and the low frequency interference are estimated through the fuzzy logic system according to a given adaptive law To the filtering back stepping controller, thereby obtaining a non-linear function required by the filtering back stepping controller. 3. The method according to claim 1 or 2,
Introducing a second filter to the filtering back stepping controller (3) in the control system (2), and approximating the virtual control amount and its derivative through a filter, the filter back stepping vessel motion control system based on the adaptive fuzzy estimator.

Images