Unmanned ship course fault-tolerant control method with input and state quantification

Abstract

The invention provides an unmanned ship course fault-tolerant control method with input and state quantification, and the method comprises the steps: considering the fault factors of a sensor and an actuator in a USV fault-tolerant system, and designing a controller; an RBF neural network is adopted to estimate uncertain items of the system; linear description is carried out on an input quantization process, so that a controller does not need priori information of any quantization parameter; the stability of the controller is proved by using the Lyapunov stability theory, and the whole closed-loop system is finally consistent and bounded. In the simulation process, a sign function in the control law is replaced with a saturation function to prevent buffeting, and the effectiveness of the strategy is verified. Fault factors such as an actuator and a sensor are considered in the ship motion control process, and fault-tolerant control over the course of the unmanned ship is achieved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an unmanned ship heading fault-tolerant control method with input and state quantization. Background Art

[0002] With the development of the shipping industry, research on unmanned vehicles (USVs) has attracted widespread attention from experts and scholars. Compared with traditional human-controlled systems, USVs are characterized by their small size, high maneuverability, and high speed. They are currently primarily used to perform tasks that are dangerous or unsuitable for human participation. Heading control, as the foundation of USV motion control, is of great research significance. During actual navigation, USV motion parameters vary with the ocean environment and the vessel's navigation conditions. Therefore, USVs exhibit highly nonlinear characteristics.

[0003] However, during the control process, the system may malfunction due to faults. Furthermore, in practical navigation, communication bandwidth is limited, and control signals must be transmitted within this bandwidth. Furthermore, in practical navigation, control signals must be transmitted over communication channels. Given the limited bandwidth at sea, it is important to quantify the inputs to the unmanned vessel's heading control process. Quantization techniques can ensure that the system operates properly within the given communication bandwidth. Currently, research on both sensor and actuator fault tolerance is limited. Summary of the Invention

[0004] Based on the above-mentioned technical problems in the motion control of unmanned ships, a fault-tolerant control method for unmanned ship heading with input and state quantization is provided. The present invention takes fault factors into consideration in the process of ship motion control and uses quantization technology to process the control input of the system.

[0005] The technical means adopted in the present invention are as follows:

[0006] A fault-tolerant control method for an unmanned ship heading with input and state quantization, comprising:

[0007] S1. Consider the failure factors of sensors and actuators in the USV fault-tolerant system and design the controller;

[0008] S2, using RBF neural network to estimate the uncertainty of the system;

[0009] S3. Linearly describe the input quantization process so that the controller does not require any prior information on the quantization parameters;

[0010] S4. Use Lyapunov stability theory to prove the stability of the controller and that the entire closed-loop system is eventually uniformly bounded.

[0011] Furthermore, step S1 specifically includes:

[0012] S11. Construct a mathematical model of the ship:

[0013]

[0014] in, is the ship heading angle, is the ship's bow angular velocity; δ is the ship's rudder angle, θ1r+θ2r 3 is the internal disturbance of the system, θ1=1 / K, θ2=α / K are known constants, K is the ship's turning index, and T is the ship's following index; d(t) is the disturbance added to the input, |d(t)|≤D;

[0015] S12. Define x1 = ψ, δ=u, g=K / T, f(x)=-K / T(θ1r+θ2r 3 ),

[0016] Q(u) is the quantized control input of the system, so the mathematical model of ship heading control after input quantization is introduced as follows:

[0017]

[0018] S13. Based on the fault factors, consider the fault tolerance of sensors and actuators and design the fault-tolerant subsystem as follows:

[0019]

[0020] Where 0<ρ i0 ≤ρ i ≤1,ρ i0 and ρ i is an unknown constant;

[0021] S14. Considering the partial failure of the actuator, the actual control input is u=ρ0Q(u),ρ0∈(0,1), and the sensor measured output is and

[0022] Furthermore, step S2 specifically includes:

[0023] S21. Approximation of the unknown function f(x) is achieved by introducing the RBF neural network. The algorithm of the RBF neural network is:

[0024]

[0025] f(x)=W *T h(x)+∈∈

[0026] Among them, x is the input of the network, i is the number of inputs of the network, j is the jth node of the hidden layer of the network, h=[h1,h2,…,h n ] T is the output of the Gaussian function, W * is the ideal weight of the network, ∈ is the approximation error of the network, |∈|≤∈ N ;

[0027] S22, using the RBF neural network to approximate the unknown function f(x), the input of the RBF neural network is x, and the output of the RBF neural network is:

[0028]

[0029] in,

[0030] Furthermore, the above-mentioned fault-tolerant control method for the heading of an unmanned ship with input and state quantization is characterized in that step S3 specifically includes:

[0031] Let Q(u)=q1(t)u+q2(t), then take:

[0032]

[0033] Among them, q1(t) is an unknown parameter. Since the sign remains unchanged during the quantization process, it can be concluded from the above formula that q1(t)>0. Since when |u(t)|<a, Q(u) is bounded, q1(t)=1, and thus q2(t) is bounded.

[0034] Furthermore, step S4 specifically includes:

[0035] S41. Definition and design Where c>0, taking the derivative of s, we have:

[0036]

[0037] Let θ=gρ1ρ0q1, η=gρ1ρ0, then we have:

[0038]

[0039] Pick If l>0, we get:

[0040]

[0041] According to the above analysis, since the parameter θ is an unknown term, the parameter θ is estimated; for the adaptive estimation of the time-varying parameter θ=gρ1ρ0q1, since the parameter cannot be differentiated, its bound needs to be estimated; for the time-varying gain θ=gρ1ρ0q1, take where θ min is the lower bound of θ;

[0042] S42. Design the Lyapunov function as follows:

[0043]

[0044] Where γ>0, From θ>0, we know that μ>0;

[0045] S43. Introduce RBF neural network to realize the approximation of unknown function f(x) and design Lyapunov function as follows:

[0046]

[0047] Where γ1>0, and taking the derivative of the Lyapunov function V, we get:

[0048]

[0049] S44. Design the control law and adaptive law as follows:

[0050]

[0051]

[0052] Among them, ρ>0, α>0, φ≥∈ n +D;

[0053] S45. From the Lyapunov function set in the controller design process, we know:

[0054]

[0055] because Then there is Pick Then there is Taking into account Then there is We further learn that:

[0056]

[0057] S46. Substituting the designed adaptive rate into the above formula, we obtain:

[0058]

[0059] because in, Then we can conclude Thus we can conclude that:

[0060]

[0061] in, λ=min{2l,γα}.

[0062] S47. Calculate inequality equations The solution is:

[0063]

[0064] Take the limit Therefore, the closed-loop system error signal is bounded and the convergence rate depends on λ and d.

[0065] Compared with the prior art, the present invention has the following advantages:

[0066] 1. The present invention provides an unmanned ship heading fault-tolerant control method with input and state quantization, which takes failure factors such as actuators and sensors into account in the ship motion control process, thereby realizing fault-tolerant control of the unmanned ship heading.

[0067] 2. The present invention provides an unmanned ship heading fault-tolerant control method with input and state quantization, which uses quantization technology to process the control input of the system, so that the controller does not need to obtain any prior information of the quantization parameters.

[0068] 3. The present invention provides an unmanned ship heading fault-tolerant control method with input and state quantization, which uses RBF neural network to estimate the uncertainties in the system. Through simulation experiments, the effectiveness of the neural network ship heading fault-tolerant control strategy with input quantization is verified.

[0069] Based on the above reasons, the present invention can be widely promoted in fields such as artificial intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0071] Figure 1 Flow chart of the method of the present invention.

[0072] Figure 2This is a diagram of the heading angle tracking results provided by an embodiment of the present invention.

[0073] Figure 3 This is the yaw angular velocity tracking result provided by the embodiment of the present invention.

[0074] Figure 4 This is a control input diagram before quantization provided by an embodiment of the present invention.

[0075] Figure 5 This is a quantized control input diagram provided by an embodiment of the present invention.

[0076] Figure 6 This is a diagram of the heading angle tracking error provided by an embodiment of the present invention.

[0077] Figure 7 This is a diagram of the yaw angular velocity tracking error provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0078] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0079] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products or apparatuses.

[0080] like Figure 1 As shown, the present invention provides an unmanned ship heading fault-tolerant control method with input and state quantization, comprising:

[0081] S1. Consider the failure factors of sensors and actuators in the USV fault-tolerant system and design the controller;

[0082] S2, using RBF neural network to estimate the uncertainty of the system;

[0083] S3. Linearly describe the input quantization process so that the controller does not require any prior information on the quantization parameters;

[0084] S4. Use Lyapunov stability theory to prove the stability of the controller and that the entire closed-loop system is eventually uniformly bounded.

[0085] In specific implementation, as a preferred embodiment of the present invention, step S1 specifically includes:

[0086] S11. Construct a mathematical model of the ship:

[0087]

[0088] in, is the ship heading angle, is the ship's bow angular velocity; δ is the ship's rudder angle, θ1r+θ2r 3 is the internal disturbance of the system, θ1=1 / K, θ2=α / K are known constants, K is the ship's turning index, and T is the ship's following index; d(t) is the disturbance added to the input, |d(t)|≤D;

[0089] S12. Define x1 = ψ, δ=u, g=K / T, f(x)=-K / T(θ1r+θ2r 3 ), Q(u) is the quantized control input of the system, so the mathematical model of ship heading control after input quantization is introduced as follows:

[0090]

[0091] S13. Based on the fault factors, consider the fault tolerance of sensors and actuators and design the fault-tolerant subsystem as follows:

[0092]

[0093] Where 0<ρ i0 ≤ρ i ≤1,ρ i0 and ρ i is an unknown constant;

[0094] S14. Considering the partial failure of the actuator, the actual control input is u=ρ0Q(u),ρ0∈(0,1), and the sensor measured output is and

[0095] In specific implementation, as a preferred embodiment of the present invention, step S2 specifically includes:

[0096] S21. Approximation of the unknown function f(x) is achieved by introducing the RBF neural network. The algorithm of the RBF neural network is:

[0097]

[0098] f(x)=W *T h(x)+∈∈

[0099] Among them, x is the input of the network, i is the number of inputs of the network, j is the jth node of the hidden layer of the network, h=[h1,h2,…,h n ] T is the output of the Gaussian function, W * is the ideal weight of the network, ∈ is the approximation error of the network, |∈|≤∈ N ;

[0100] S22, using the RBF neural network to approximate the unknown function f(x), the input of the RBF neural network is x, and the output of the RBF neural network is:

[0101]

[0102] in,

[0103] In specific implementation, as a preferred embodiment of the present invention, step S3 specifically includes:

[0104] Let Q(u)=q1(t)u+q2(t), then take:

[0105]

[0106] Among them, q1(t) is an unknown parameter. Since the sign remains unchanged during the quantization process, it can be concluded from the above formula that q1(t)>0. Since when |u(t)|<a, Q(u) is bounded, q1(t)=1, and thus q2(t) is bounded.

[0107] In specific implementation, as a preferred embodiment of the present invention, step S4 specifically includes:

[0108] S41. Definition and design Where c>0, taking the derivative of s, we have:

[0109]

[0110] Let θ=gρ1ρ0q1, η=gρ1ρ0, then we have:

[0111]

[0112] Pick If l>0, we get:

[0113]

[0114] According to the above analysis, since the parameter θ is an unknown term, the parameter θ is estimated; for the adaptive estimation of the time-varying parameter θ=gρ1ρ0q1, since the parameter cannot be differentiated, its bound needs to be estimated; for the time-varying gain θ=gρ1ρ0q1, take where θ min is the lower bound of θ;

[0115] S42. Design the Lyapunov function as follows:

[0116]

[0117] Where γ>0, From θ>0, we know that μ>0;

[0118] S43. Introduce RBF neural network to realize the approximation of unknown function f(x) and design Lyapunov function as follows:

[0119]

[0120] Where γ1>0, and taking the derivative of the Lyapunov function V, we get:

[0121]

[0122] S44. Design the control law and adaptive law as follows:

[0123]

[0124] Among them, ρ>0, α>0, φ≥∈ n +D;

[0125] S45. From the Lyapunov function set in the controller design process, we know:

[0126]

[0127] because Then there is Pick Then there is Taking into account Then there is We further learn that:

[0128]

[0129] S46. Substituting the designed adaptive rate into the above formula, we obtain:

[0130]

[0131] because in, Then we can conclude Thus we can conclude that:

[0132]

[0133] in, λ=min{2l,γα}.

[0134] S47. Calculate inequality equations The solution is:

[0135]

[0136] Take the limit Therefore, the closed-loop system error signal is bounded and the convergence rate depends on λ and d.

[0137] Example

[0138] To verify the effectiveness of the present invention, this embodiment uses MATLAB to conduct simulation research, and the parameters are set as follows:

[0139] Take the Dalian Maritime University's "Lanxin" unmanned vessel as an example: the unmanned vessel is 7.02 meters long, 2.60 meters wide, has a fully loaded draft of 0.32 meters, and a block coefficient of 0.6976. The vessel's model parameters are K = 0.71, T = 0.32, and the Norrbin motion model is used with a nonlinear coefficient a = 0.001. Set the ideal heading command y d = sint, d(t) = sin(π*t). The controller parameter model is set as: c = 15, l = 15, ρ1 = 0.95, ρ2 = 0.95, α = 0.2, c1 = 2, ρ = 0.02, γ = 2, γ1 = 1, D = 1, φ = D + 0.1 = 1.1, k = 2.

[0140] In order to prevent chattering, the saturation function sat(s) is used in the controller instead of the sign function sgn(s), that is:

[0141]

[0142] Among them, Δ is the boundary layer, Δ=0.05, RBF neural network is used, the network structure is 1-5-1, and the parameters of the Gaussian basis function are designed according to the actual range of the network input. The parameter c i and b i The values ​​are [-2-1 0 1 2] and 3.0 respectively. The initial value of each element in the network weight is 0.10.

[0143] The simulation results are as follows Figure 2-Figure 7 As shown, Figure 2 is the heading angle tracking result diagram, Figure 3 is the bow angular velocity tracking result diagram, Figure 4 is the control input map before quantization, Figure 5 is the quantized control input map, Figure 6 is the heading angle tracking error diagram, Figure 7 This is the bow angular velocity tracking error diagram. It can be seen from the above diagram that the unmanned ship heading fault-tolerant control method with input and state quantization provided by the present invention will not reduce the stability of the control system, and further verifies the effectiveness of the proposed strategy.

[0144] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fault-tolerant control method for unmanned ship heading with input and state quantization, characterized in that: include: S1. Consider the failure factors of sensors and actuators in the USV fault-tolerant system and design the controller; S2, using RBF neural network to estimate the uncertainty of the system; S3. Linearly describe the input quantization process so that the controller does not require any prior information on the quantization parameters; S4. Use Lyapunov stability theory to prove the stability of the controller and that the entire closed-loop system is eventually uniformly bounded.

2. The method for fault-tolerant heading control of an unmanned ship with input and state quantization according to claim 1, characterized in that: Step S1 specifically includes: S11. Construct a mathematical model of the ship: in, is the ship heading angle, is the ship's bow angular velocity; δ is the ship's rudder angle, θ1r+θ2r 3 is the internal disturbance of the system, θ1=1 / K, θ2=α / K are known constants, K is the ship's turning index, and T is the ship's following index; d(t) is the disturbance added to the input, |d(t)|≤D; S12、definingx1=ψ, δ=u, g=K / T, f(x)=-K / T(θ1r+θ2r 3 ), Q(u) is the quantized control input of the system, so the mathematical model of ship heading control after input quantization is introduced as follows: S13. Based on the fault factors, consider the fault tolerance of sensors and actuators and design the fault-tolerant subsystem as follows: Where 0<ρ i0 ≤ρ i ≤1,ρ i0 and ρ i is an unknown constant; S14. Considering the partial failure of the actuator, the actual control input is u=ρ0Q(u),ρ0∈(0,1), and the sensor measured output is and 3. The method for fault-tolerant heading control of an unmanned ship with input and state quantization according to claim 1, characterized in that: Step S2 specifically includes: S21. Approximation of the unknown function f(x) is achieved by introducing the RBF neural network. The algorithm of the RBF neural network is: f(x)=W *T h(x)+ Among them, x is the input of the network, i is the number of inputs of the network, j is the jth node of the hidden layer of the network, h=[h1,h2,…,h n ] T is the output of the Gaussian function, W * is the ideal weight of the network, ∈ is the approximation error of the network, |∈|≤∈ N ; S22, using the RBF neural network to approximate the unknown function f(x), the input of the RBF neural network is x, and the output of the RBF neural network is: in, 4. The method for fault-tolerant control of unmanned ship heading with input and state quantization according to claim 1, characterized in that: Step S3 specifically includes: Let Q(u)=q1(t)u+q2(t), then take: Among them, q1(t) is an unknown parameter. Since the sign remains unchanged during the quantization process, it can be concluded from the above formula that q1(t)>0. Since when |u(t)|<a, Q(u) is bounded, q1(t)=1, and thus q2(t) is bounded.

5. The method for fault-tolerant heading control of an unmanned ship with input and state quantization according to claim 1, characterized in that: Step S4 specifically includes: S41. Definition and design Where c>0, taking the derivative of s, we have: Let θ=gρ1ρ0q1, η=gρ1ρ0, then we have: Pick Then we get: According to the above analysis, since the parameter θ is an unknown term, the parameter θ is estimated; for the adaptive estimation of the time-varying parameter θ=gρ1ρ0q1, since the parameter cannot be differentiated, its bound needs to be estimated; for the time-varying gain θ=gρ1ρ0q1, take where θ min is the lower bound of θ; S42. Design the Lyapunov function as follows: Where γ>0, From θ>0, we know that μ>0; S43. Introduce RBF neural network to realize the approximation of unknown function f(x) and design Lyapunov function as follows: Where γ1>0, and taking the derivative of the Lyapunov function V, we get: S44. Design the control law and adaptive law as follows: Where, ρ>0, α>0, φ≥∈ n +D; S45. From the Lyapunov function set in the controller design process, we know: because Then there is Pick Then there is Taking into account Then there is We further learn that: S46. Substituting the designed adaptive rate into the above formula, we obtain: because in, Then we can conclude Thus we can conclude that: Among them, λ=min{2l,γα}. S47. Calculate inequality equations The solution is: Take the limit Therefore, the closed-loop system error signal is bounded and the convergence rate depends on λ and d.

Images