CN110989563A - Fault estimation method for unmanned ships based on iterative adaptive observer - Google Patents

Abstract

The invention relates to an iterative adaptive observer-based fault estimation method for an unmanned ship, belonging to the technical field of unmanned ship control. system, in which subsystem 1 only contains servo faults, and subsystem 2 only contains sensor faults; for subsystem 1, an adaptive fault observer is designed to estimate the efficiency factor of the steering gear; for subsystem 2, an iterative adaptive fault observer is designed to estimate Sensor failure; establish the error equation of subsystem 1 and subsystem 2 to judge the stability of the error system. The invention can realize the accurate estimation of the failure of the unmanned boat system, and provide information such as the time of occurrence of the failure, the development process and the severity of the failure, which is convenient for the operation center to monitor the safety of the unmanned boat; the invention can also Estimate the failure of the steering gear and the sensor of the unmanned boat at the same time, which reduces the cost of fault-tolerant design.

Description

Fault estimation method for unmanned ships based on iterative adaptive observer

technical field

The invention relates to an unmanned ship fault estimation method based on an iterative adaptive observer, and belongs to the technical field of unmanned ship control.

Background technique

As an autonomous motion platform working in a complex marine environment, unmanned ships can perform tasks in high-risk sea areas for a long time due to their unmanned and intelligent operation and no casualties. At the same time, unmanned ships are small, flexible, and highly concealed. By loading diversified equipment for different mission requirements, they can play an important role in territorial sea cruises, intelligence reconnaissance, and marine resource exploration.

On the other hand, due to the complex marine environment and harsh climatic conditions, when unmanned surface ships perform scientific research and exploration tasks for a long time, various components such as steering gears and sensors will inevitably fail, which will not only cause unmanned ships. The degradation of system performance can even damage the high-precision detection equipment carried on the hull. Therefore, it is very important to visually display the amplitude and frequency of unmanned boat faults in real time through fault estimation technology, and to effectively monitor the health status of unmanned boats, which is of great importance to improve the safety and reliability of unmanned boat systems. significance.

The problem of fault diagnosis and fault-tolerant design of unmanned ships has always been a research hotspot in the field of control. However, the existing methods mainly have the following two problems: First, the existing surface craft fault diagnosis methods often use the residual evaluation function generated by the fault filter to qualitatively judge whether there is a fault, but it cannot provide accurate fault diagnosis. Quantitative information, such as fault amplitude, shape, etc., makes it impossible for the operation center to judge the time, development process and severity of the failure of the unmanned boat system, which brings certain benefits to the compensation of subsequent failures and the design of fault tolerance strategies. difficulty. Secondly, the current UAV fault diagnosis methods often assume that there is only a steering gear fault on the ship, or only diagnose sensor faults. However, the components that fail during actual operation are uncertain, so multiple diagnostic modules often need to be designed on unmanned boats, which increases the cost of fault-tolerant design.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide an iterative adaptive observer-based unmanned ship fault estimation method in order to solve the problem that the existing unmanned ship fault diagnosis technology cannot provide quantitative information of faults and requires multiple groups of modules to monitor the ship's faults. .

The purpose of the present invention is to achieve this: an iterative adaptive observer-based fault estimation method for unmanned ships, which specifically includes the following steps:

Step 1. Decompose the unmanned surface vehicle model containing both the steering gear failure and the sensor fault into two subsystems through coordinate transformation, wherein the subsystem 1 only contains the steering gear fault, and the subsystem 2 only contains the sensor fault;

Step 2. For subsystem 1 in step 1, design an adaptive fault observer to estimate the efficiency factor of the steering gear;

Step 3. For subsystem 2 in step 1, design an iterative adaptive fault observer to estimate sensor faults;

Step 4, establish the error equation of subsystem 1 and subsystem 2, and judge the stability of the error system.

The present invention also includes such structural features:

1. The step 1 specifically includes the following steps:

Step 1.1. Establish a mathematical model of the unmanned surface vehicle with steering gear failure and sensor failure:

y _o (t)=C _o x _o (t)+D _os f _s (t)

Among them, the state vector x _o (t)=[v(t)r(t)ψ(t)p(t)φ(t)] ^T , the disturbance vector d(t)=[w _ψ w _φ ] ^T , the control Input u(t)=δ(t), ρ(t) is the unknown actuator efficiency factor, y _o (t)∈R ^p is the measured output signal, f _s (t)=[f _s1 ,f _s2 , ...,f _sq ] ^T ∈ R ^q is the sensor fault vector; the variables v(t), ψ(t), φ(t), r(t) and p(t) represent the rudder-generated lateral moving speed, heading angle, roll angle, yaw speed and roll speed; δ(t) represents the rudder angle; w _φ (t) and w _ψ (t) represent the disturbance of roll angle and heading angle caused by sea waves; T _v and T _r are time constants; K _vr , K _dv , K _dv , K _dr , K _dp , K _vp are known gains; _wn and ζ represent the undamped natural frequency and damping ratio, respectively, and the system matrix is expressed as :

Co＝I₅

Co=I ₅

Other model parameters, disturbances and faults must satisfy: Do _os ∈ R ^5×q is the sensor fault coefficient matrix, which satisfies q≤5;

d*，s_uv，s_dv为未知的正数；The efficiency factor ρ(t) of the steering gear satisfies 0<ρ(t)≤ρ _u ≤1; the disturbance d(t) satisfies ||d(t)||≤d*; the sensor fault f _sv (t) satisfies For |f _sv (t)|≤s _uv ,

d*, s _uv , s _dv are unknown positive numbers;

Step 1.2. For the original system 1, find the reversible matrix M∈R ^5×5 and N∈R ^5×5 , satisfying:

Among them, A ₁ ∈ R ^3×3 , G ₁ ∈ R ^3×1 , G ₂ ∈ R ^3×2 , C ₁ ∈ R ^3×3 are reversible, D∈R ^2×q ;

x₁(t)∈R³，y₁(t)∈R³，根据步骤1.2中公式,可以分别得到只含有舵机故障与传感器故障的两个子系统：Step 1.3. Introduce linear transformation x(t)=Mx _o (t), y(t)=Ny _o (t),

x ₁ (t)∈R ³ , y ₁ (t)∈R ³ , according to the formula in step 1.2, two subsystems containing only servo faults and sensor faults can be obtained respectively:

Step 1.4. The augmentation processing process for the sensor fault formula in step 1.3 is as follows:

First, define a measurable output variable

给定

则步骤1.4中舵机故障公式可被重写为：Then, construct the augmented vector

given

Then the servo fault formula in step 1.4 can be rewritten as:

Subsystem 1:

Finally, according to the sensor failure formula in step 1.3 and the first formula in step 1.4, the augmented subsystem 2 can be obtained:

其中，ξ(t)∈R²是子系统2的量测输出。Subsystem 2:

where ξ(t)∈R ² is the measurement output of subsystem 2.

2. The step 2 specifically includes: designing an adaptive observer for subsystem 1 in the subsystem formula in step 1.4 as follows:

是x₁(t)的估计值；

是

在第k个迭代观测器中的估计值；

是执行器效率因子ρ(t)的估计值；A_f∈R^3×3是待设计的矩阵参数，将在步骤4.2矩阵A_f求解公式中给出；观测器输入u_d(t)设计为：in,

is an estimate of x ₁ (t);

Yes

the estimated value in the k-th iteration observer;

is the estimated value of the actuator efficiency factor ρ(t); A _f ∈ R ^3×3 is the matrix parameter to be designed, which will be given in the solution formula of matrix A _f in step 4.2; the observer input u _d (t) is designed as :

是d^*的自适应估计；P∈R^3×3是待设计的正定对称矩阵，将在步骤4.2第一个公式中给出；

b₀和b₁是给定的正数；in,

is the adaptive estimate of d ^* ; P ∈ R ^3×3 is the positive definite symmetric matrix to be designed, which will be given in the first formula of step 4.2;

b ₀ and b ₁ are given positive numbers;

和

的自适应律设计为：

and

The adaptive law is designed as:

where c ₁ , c _d1 , and c _d2 are given positive numbers.

3. The step 3 specifically includes: designing an iterative adaptive observer for subsystem 2 in the formula of subsystem 2 in step 1.4 is:

是

的第k个迭代观测器中的估计值。θ是迭代次数最大值。L_p∈R^4×2是观测器增益满足于

是一个稳定的矩阵；in,

Yes

The estimated value in the k-th iteration observer of . θ is the maximum number of iterations. L _p ∈ R ^4×2 is the observer gain satisfying

is a stable matrix;

是传感器故障的估计值，其迭代自适应率为：k＝1时：

is the estimated value of sensor failure, and its iterative adaptation rate is: when k=1:

When k≥2:

where c _sv is a given positive number; Π _1v is the vth row of the matrix Π ₁ ∈ R ^q×2 , Π _2v is the vth row of the matrix Π ₂ ∈ R ^q×2 , v=1,2,. ..,q, the matrices Π ₁ and Π ₂ are calculated by the first formula in step 4.2.

4. The stability of the step 4 judgment error system specifically includes the following steps:

Step 4.1. Define the estimation error as:

The dynamic equation of the error system is established as:

Step 4.2. Judge the stability of the above formula according to the following conditions:

Π₁∈R^q×2，Π₂∈R^q×2，满足：If there are positive constants δ _s , γ, positive definite matrices P∈R ^3×3 , Q∈R ^4×4 , the matrix

Π ₁ ∈ R ^q×2 , Π ₂ ∈ R ^q×2 , satisfy:

in

Then the dynamic equation of the error system is bounded and stable, the parameter matrix P∈R ^3×3 , the matrix Π ₁ ∈R ^q×2 , Π ₂ ∈R ^q×2 can be obtained by the first formula in step 4.2, the matrix A _f can be obtained by solving Solve as:

Other adaptive parameters c ₁ , c _d1 , c _d2 , c _s1 , b ₀ , b ₁ in the formulas of the detailed steps of step 2 and step 3 are given free positive numbers, which can be adjusted according to the effect of the observer.

Compared with the prior art, the beneficial effects of the present invention are: the present invention can provide quantitative information such as the amplitude and shape of the unmanned boat fault in real time, and intuitively display the development process of the fault and the severity of the fault, which is convenient for control. The center monitors the safety of unmanned boats to improve the reliability of unmanned boats in the process of performing tasks. The present invention can also perform unified monitoring on the failure situation of the steering gear of the unmanned boat and the fault of the sensor, which reduces the construction of the fault diagnosis module on the unmanned boat, and effectively reduces the cost of fault-tolerant design.

Description of drawings

Fig. 1 is the flow chart of the present invention;

Fig. 2 is the system block diagram of the present invention;

3 is a schematic diagram of the actual efficiency factor (solid line) of the steering gear and the estimated value (dotted line) of the steering gear efficiency factor in the simulation example;

Figure 4 is a schematic diagram of the real sensor failure (solid line) and the estimated sensor failure (dashed line) in the simulation example.

Detailed ways

The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

The present invention provides a fault estimation method for an unmanned surface vessel based on an iterative adaptive observer, as shown in FIG. 1 , and the specific implementation steps are as follows:

Step 1: Decompose the unmanned surface vehicle model containing both the steering gear failure and the sensor failure into two subsystems through coordinate transformation, wherein the subsystem 1 only contains the steering gear fault, and the subsystem 2 only contains the sensor fault. The step 1 includes: :

Step A. According to the document "Integral-Based Event-Triggered Fault Detection Filter Design for Unmanned Surface Vehicles", the mathematical model of the unmanned surface vehicle with steering gear failure and sensor failure is established as:

y _o (t)=C _o x _o (t)+D _os f _s (t) (1)

Among them, the state vector x _o (t)=[v(t)r(t)ψ(t)p(t)φ(t)] ^T , the disturbance vector d(t)=[w _ψ w _φ ] ^T , the control Input u(t)=δ(t), ρ(t) is the unknown actuator efficiency factor, y _o (t)∈R ^p is the measured output signal, f _s (t)=[f _s1 ,f _s2 , ...,f _sq ] ^T ∈ R ^q is the sensor fault vector; the variables v(t), ψ(t), φ(t), r(t) and p(t) represent the rudder-generated lateral moving speed, heading angle, roll angle, yaw speed and roll speed; δ(t) represents the rudder angle; w _φ (t) and w _ψ (t) represent the disturbance of roll angle and heading angle caused by sea waves; T _v and T _r are time constants; K _vr , K _dv , K _dv , K _dr , K _dp , K _vp are known gains; _wn and ζ represent the undamped natural frequency and damping ratio, respectively, and the system matrix is expressed as :

Given parameters: K _vr = -0.46m/s, ω _n =0.63rad/s, ζ = 0.0936, K _dr = -0.0211, T _r =1.6667, T _v =10, K _dv =0.0780, K _vp =1.6380 , K _dp =-0.0852, the parameter matrix can be obtained:

B _o = [0.0078 -0.0126 0 -0.0338 0] ^T ,

C_o＝I₅

C _o =I ₅

The sensor failure coefficient matrix is assumed to be _Dos = [0 0 1 0 1] ^T .

Step B. For the original system (1), obtain the invertible matrix

和

满足于：

and

satisfied with:

x₁(t)∈R³，y₁(t)∈R³，分别只含有舵机故障与传感器故障的两个子系统可以表示为：Step C, according to formula (3), introduce linear transformation x(t)=Mx _o (t), y(t)=Ny _o (t), wherein

x ₁ (t)∈R ³ , y ₁ (t)∈R ³ , the two subsystems containing only servo fault and sensor fault respectively can be expressed as:

Step D, the process of augmenting the sensor fault in the formula (5) is:

First, define a measurable output variable

给定

则式(4)可被重写为：Then, construct the augmented vector

given

The formula (4) can be rewritten as:

Subsystem 1:

Finally, according to equations (5) and (6), the augmented subsystem 2 can be obtained:

其中，ξ(t)∈R²是子系统2的量测输出。Subsystem 2:

where ξ(t)∈R ² is the measurement output of subsystem 2.

Step 2: Design an adaptive fault observer to estimate the efficiency factor of the steering gear for the subsystem 1 in the step 1, and the step 2 includes:

Step E. Design an adaptive observer for subsystem 1 in equation (7) as:

是x₁(t)的估计值；

是

在第k个迭代观测器中的估计值；

是执行器效率因子ρ(t)的估计值；A_f∈R^3×3是待设计的矩阵参数，将在式(20)中给出；观测器输入u_d(t)设计为：in,

is an estimate of x ₁ (t);

Yes

the estimated value in the k-th iteration observer;

is the estimated value of the actuator efficiency factor ρ(t); A _f ∈ R ^3×3 is the matrix parameter to be designed, which will be given in equation (20); the observer input u _d (t) is designed as:

是d^*的自适应估计；P∈R^3×3是待设计的正定对称矩阵，将在(18)中给出；

b₀和b₁是给定的正数。in,

is the adaptive estimate of d ^* ; P ∈ R ^3×3 is the positive definite symmetric matrix to be designed, which will be given in (18);

b ₀ and b ₁ are given positive numbers.

和

的自适应律设计如下：

and

The adaptive law is designed as follows:

where c ₁ , c _d1 , and c _d2 are given positive numbers.

Step 3: Design an iterative adaptive fault observer to estimate sensor faults for subsystem 2 in Step 1. Step 3 includes:

Step F. For subsystem 2 in equation (8), design an iterative adaptive observer as follows:

是

的第k个迭代观测器中的估计值。θ是迭代次数最大值。

是观测器增益满足于

是一个稳定的矩阵；in,

Yes

The estimated value in the k-th iteration observer of . θ is the maximum number of iterations.

is the observer gain satisfying

is a stable matrix;

是传感器故障的估计值，其迭代自适应率为：k＝1时：

is the estimated value of sensor failure, and its iterative adaptation rate is: when k=1:

When k≥2:

where c _sv is a positive number; Π _1v is the vth row of the matrix Π ₁ ∈ R ^q×2 , Π _2v is the vth row of the matrix Π ₂ ∈ R ^q×2 , v=1,2,... , q, matrices Π ₁ and Π ₂ are calculated by formula (18).

Step 4: Establish the error equations of subsystem 1 and subsystem 2, and give the stability conditions of the error system. The step 4 includes:

First, define the estimation error as:

The dynamic equation of the error system is established as:

The stability of the error equation (17) is judged according to the following conditions:

Π₁∈R^q×2，Π₂∈R^q×2，满足于If there are positive constants δ _s , γ, positive definite matrices P∈R ^3×3 , Q∈R ^4×4 , the matrix

Π ₁ ∈ R ^q×2 , Π ₂ ∈ R ^q×2 , satisfy

in

Then the dynamic error (17) is bounded and stable, the parameter matrix P∈R ^3×3 , Π ₁ ∈ R ^q×2 , Π ₂ ∈ R ^q×2 can be obtained by solving (18), and each parameter matrix is solved as:

Other adaptive parameters in equations (10)-(12), (14)-(15) are given as c ₁ =1000, c _d1 =1, c _d2 =1, c _s1 =2.1, b ₀ =0.01 , b ₁ =0.01.

同时根据迭代自适应观测器(13)与迭代更新率(14)-(15)，我们可以获得第k次迭代过程的传感器故障f_s(t)的估计值

Through the above design process, based on the adaptive observer (9) and the update rate (11), we can obtain the estimated value of the efficiency factor ρ(t) of the steering gear

At the same time, according to the iterative adaptive observer (13) and the iterative update rate (14)-(15), we can obtain the estimated value of the sensor failure f _s (t) of the k-th iteration process

In order to check the effect of the present invention, the following simulation examples are used for verification. Suppose the failure of the steering gear and the sensor failure are:

ρ(t)=0.2e ^-0.2t +0.35

The disturbance w _ψ (t) of the heading angle caused by the sea wave is a random value in [-1,1], the disturbance of the roll angle caused by the sea wave w _φ (t)=0.5sin(1.2t), the control input u(t) =0.1 sin(t).

The failure situation of the steering gear and the estimation results of the observer are shown in Figure 3, where the solid line is the real curve of the efficiency factor of the steering gear, and the dotted line is the estimated value of the efficiency factor of the steering gear obtained by the observer. The sensor failure and observer estimation results are shown in Figure 4, where the solid line is the real curve of the sensor failure, and the dashed line is the estimated value of the sensor failure obtained by the observer.

It can be concluded from the simulation results that the present invention can estimate the efficiency factor of the steering gear of the unmanned surface craft and the fault of the sensor at the same time, and can quantitatively describe the occurrence of the fault quickly and accurately, which is convenient for the control center to ensure the safety of the unmanned craft. Sexual surveillance.

The descriptions not involved in the specific embodiments of the present invention belong to the known technology in the art, and can be implemented with reference to the known technology. The present invention is not limited to this specific embodiment, and any modification or equivalent replacement of the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention shall be included in the protection scope of the present invention.

To sum up, the present invention discloses an iterative adaptive observer-based fault estimation method for an unmanned surface vehicle. system; step 2, designing an adaptive fault observer to estimate the efficiency factor of the steering gear; step 3, designing an iterative adaptive fault observer to estimate the sensor fault; and step 4, judging the stability of the error system. The invention can realize the accurate estimation of the failure of the unmanned boat system, and provide information such as the time of occurrence of the failure, the development process and the severity of the failure, which is convenient for the operation center to monitor the safety of the unmanned boat; The failure of the steering gear of the unmanned boat is estimated at the same time as the sensor fault, which avoids the construction of multiple fault diagnosis modules on the unmanned boat, and reduces the cost of fault-tolerant design.

Claims (1)

1. an unmanned ship fault estimation method based on iterative adaptive observer, is characterized in that, specifically comprises the following steps: Step 1: Decompose the unmanned surface vehicle model containing both the steering gear failure and the sensor fault into two subsystems through coordinate transformation, wherein the subsystem 1 only contains the steering gear fault, and the subsystem 2 only contains the sensor fault; Step 1.1: Build the mathematical model of the UUV with steering gear failure and sensor failure:

y _o (t)=C _o x _o (t)+D _os f _s (t) Among them, the state vector x _o (t)=[v(t) r(t) ψ(t) p(t) φ(t)] ^T , the disturbance vector d(t)=[w _ψ w _φ ] ^T , the control Input u(t)=δ(t), ρ(t) is the unknown actuator efficiency factor, y _o (t)∈R ^p is the measured output signal, f _s (t)=[f _s1 ,f _s2 , ...,f _sq ] ^T ∈ R ^q is the sensor fault vector; the variables v(t), ψ(t), φ(t), r(t) and p(t) represent the rudder-generated lateral moving speed, heading angle, roll angle, yaw speed and roll speed; δ(t) represents the rudder angle; w _φ (t) and w _ψ (t) represent the disturbance of roll angle and heading angle caused by sea waves; T _v and T _r are time constants; K _vr , K _dv , K _dv , K _dr , K _dp , K _vp are known gains; _wn and ζ represent the undamped natural frequency and damping ratio, respectively, and the system matrix is expressed as :

C _o =I ₅ Other model parameters: interference and fault must satisfy: Do _os ∈ R ^5×q is the sensor fault coefficient matrix, satisfying q≤5; The efficiency factor ρ(t) of the steering gear satisfies 0<ρ(t)≤ρ _u ≤1; the disturbance d(t) satisfies ||d(t)||≤d ^* ; the sensor fault f _sv (t) satisfies For |f _sv (t)|≤s _uv ,

v=1,2,...,q; d ^* , s _uv , s _dv are unknown positive numbers; Step 1.2: For the original system 1, find the reversible matrices M∈R ^5×5 and N∈R ^5×5 , satisfying:

Among them, A ₁ ∈ R ^3×3 , G ₁ ∈ R ^3×1 , G ₂ ∈ R ^3×2 , C ₁ ∈ R ^3×3 are reversible, D∈R ^2×q ; Step 1.3: Introduce linear transformation x(t)=Mx _o (t), y(t)=Ny _o (t),

x ₁ (t)∈R ³ , y ₁ (t)∈R ³ , two subsystems containing only servo faults and sensor faults are obtained:

Step 1.4: Augment the sensor failure formula: First, define a measurable output variable

Then, construct the augmented vector

given

Then the servo fault formula can be rewritten as: Subsystem 1:

Finally, the augmented subsystem 2 is obtained: Subsystem 2:

Among them, ξ(t)∈R ² is the measurement output of subsystem 2; Step 2: For subsystem 1, estimate the efficiency factor of the steering gear according to the adaptive fault observer;

in,

is an estimate of x ₁ (t);

Yes

the estimated value in the k-th iteration observer;

is the estimated value of the actuator efficiency factor ρ(t); A _f ∈ R ^3×3 is the matrix parameter to be calculated; the observer input u _d (t) is:

in,

is the adaptive estimate of d ^* ; P ∈ R ^3×3 is the positive definite symmetric matrix to be computed;

b ₀ and b ₁ are given positive numbers;

and

The adaptive law of is:

Among them, c ₁ , c _d1 , c _d2 are given positive numbers; Step 3: For subsystem 2, estimate the sensor fault according to the iterative adaptive fault observer;

in,

Yes

The estimated value in the k-th iteration observer of ; θ is the maximum number of iterations; L _p ∈ R ^{4 ×2} is the observer gain satisfying

is a stable matrix;

is an estimate of sensor failure with an iterative adaptation rate of: When k=1:

When k≥2:

where c _sv is a given positive number; Π _1v is the vth row of the matrix Π ₁ ∈ R ^q×2 , Π _2v is the vth row of the matrix Π ₂ ∈ R ^q×2 , v=1,2,. ..,q; Step 4: Establish the error equation of subsystem 1 and subsystem 2, and judge the stability of the error system; Step 4.1: Define the estimation error as:

The dynamic equation of the error system is established as:

Step 4.2: Judge the stability of the error system according to the following conditions: If there are positive constants δ _s , γ, positive definite matrices P∈R ^3×3 , Q∈R ^4×4 , the matrix

Π ₁ ∈ R ^q×2 , Π ₂ ∈ R ^{q ×2} , satisfy:

in

Then the dynamic equation of the error system is bounded and stable, the parameter matrix P∈R ^3×3 , the matrix A _f can be solved as:

Images