A model predictive fault-tolerant control method for unmanned ships based on tube
Through the tube-based model prediction control method, the multiplication uncertainty of the unmanned ship's dynamic positioning system under thruster failure and ocean disturbance is solved, and the failure and stability of the system are solved in traditional control strategies. Inadequate constraint handling.
Patent Information
- Application Number
- CN202211518043.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-29
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2042-11-29
AI Technical Summary
The existing unmanned ship dynamic positioning system is difficult to maintain stability when the thruster fails, and traditional model prediction control cannot effectively deal with system constraints and multiplication uncertainty problems.
Using a tube-based model prediction control method, the dynamic model of the unmanned ship DP system is constructed, the integral sliding mode surface and adaptive estimation law are designed, the equivalent sliding mode dynamics are split, and the tube containing fault uncertainty is designed. Update the tube cross-section parameters online, combining the model prediction controller and the integral sliding mode controller to ensure the stability of the system under faults and disturbances.
The system robustness and stability in the case of thruster failure and ocean disturbance is realized, conservatism is reduced, the system is effectively controlled under multiplication uncertainty, and the problem of state exceeding the scope of the terminal set is avoided.
Smart Images

Figure CN115933590B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of unmanned ship dynamic positioning control, and in particular to a tube-based unmanned ship model prediction fault-tolerant control method. Background Art
[0002] The dynamic positioning (DP) system of an unmanned marine vessel refers to a functional system that uses active thrusters to keep the vehicle at a fixed position and heading. With the increasing maturity of DP technology, it is widely installed on offshore service ships and other production platforms for drilling, mining sampling, environmental perception, anchor handling and other operations. The performance and reliability of DP systems are increasingly valued by researchers. Since the thruster is usually partially exposed on the sea surface, when it runs for a long time, sudden failures may occur, resulting in degradation of DP performance and even mission failure of the working ship. Therefore, fault-tolerant control methods have been developed to improve the resilience of the system to faults. Sliding mode control does not rely on the fault detection module. Integral sliding mode control introduces an integrator on the sliding surface so that the system state can be maintained on the sliding surface at the beginning, thereby eliminating the arrival stage and enhancing the robustness of the system. Therefore, the research on integral sliding mode fault-tolerant methods in theory and engineering applications has been relatively mature. For example, Hao LY et al. proposed a signal quantization integral sliding mode fault-tolerant control strategy based on output feedback. In existing research results, full rank decomposition and adaptive technology can be used to compensate for thruster failures and ocean disturbances. However, when a DP ship operates in a specific area, the inevitable state and input constraints must be considered. Meanwhile, the equivalent dynamics will be affected by the multiplicative uncertainty caused by the fault factors. The existing fault-tolerant methods cannot adapt well to the system constraints while achieving the expected control objectives.
[0003] On the other hand, model predictive control proposes an effective control method to deal with system constraints in control design. Robust model predictive control has become an interesting and challenging research topic. Bertsekas DP et al. introduced tube-based robust model predictive control to deal with system uncertainty. The basic idea of tube-based model predictive control is to restrict all possible state trajectories to a pre-designed tube to ensure that the constraints of all possible disturbances in the optimization problem are met. Due to the use of feedback control law, tube-based model predictive control reduces conservatism while maintaining similar computational complexity to traditional model predictive control. The tube is implemented by constructing a positive invariant set, and the tube parameters are used as decision variables in the optimization problem to achieve online design and update of the tuba section. Zheng H et al. proposed a tube-based robust DP controller that can handle system constraints, state errors, and environmental uncertainties well. Yang H et al. proposed a robust nonlinear model predictive solution method to solve disturbances, input saturation, and nonlinear problems in ship dynamics. However, a basic assumption of the above DP controller based on model predictive control is that the thruster always works properly. If the thruster fails, the controller may not be able to stabilize the unmanned ship system. The failure of the propeller can be regarded as a multiplicative uncertainty, and the tube-based model predictive control can handle this type of uncertainty. Therefore, it is feasible for the present invention to design a tube-based model predictive control method to solve the fault tolerance problem of the constrained unmanned ship. Summary of the invention
[0004] According to the above proposal, considering that traditional model predictive control cannot guarantee the stability of ships with propeller failures, and integral sliding mode control cannot handle system constraint problems, a tube-based model predictive control fault-tolerant control method is provided. Compared with the traditional model predictive control strategy, the tube-based model predictive control strategy proposed in the present invention can not only solve the problems of propeller failure and the influence of ocean disturbances, but also can handle the multiplicative uncertainty caused by the fault factor.
[0005] The technical means adopted by the present invention are as follows:
[0006] A tube-based model prediction fault-tolerant control method comprises the following steps:
[0007] S1: Construct the unmanned ship DP system dynamics model based on state and input constraints, thruster fault model and ocean disturbance;
[0008] S2: Design an integral sliding surface according to the state of the unmanned ship DP system dynamics model, and design an adaptive estimation law for the unknown parameters contained in the unmanned ship DP system dynamics model through an adaptive estimation method;
[0009] S3: Based on the definition of the integral sliding mode surface, the equivalent sliding mode dynamics with multiplicative uncertainty caused by the thruster failure factor is established;
[0010] S4: Discretize the equivalent sliding mode dynamics to obtain a discrete system, split the discrete system into a nominal part and an uncertain part, design a tube with fault uncertainty, and use the tube cross-sectional parameters as decision variables to update the size and shape of the tube cross-sectional area online, and design the state invariant set based on the terminal set of the nominal part;
[0011] S5: Design a tube-based model predictive controller and prove the feasibility of the optimization problem in the model predictive controller and the stability of the discrete system;
[0012] S6: Design an integral sliding mode fault-tolerant controller based on the model predictive controller and prove the asymptotic stability of the unmanned ship DP system.
[0013] Furthermore, the specific process of step S1 is as follows:
[0014] S11. Define the position vector of the unmanned ship as ξ(t) = [x p (t)y p (t)φ(t)] T , the velocity vector is Among them, (x p (t)y p (t)) represents the position information, φ(t) represents the heading angle, v(t) and r(t) represent the angular velocity of surge, sway and pitch respectively. Definition of reference state Error Status
[0015] S12. Establish a unified thruster failure model, including three failure modes: thruster failure, interruption and jamming. Its form is
[0016] u F (t)=ρu0(t)+σu s (t)
[0017] Where ρ is a positive definite diagonal matrix, representing the failure factor, and the range of each element is [0,1]; σ is the stuck failure factor, and its value is 0 or 1; u s (t) indicates a stuck fault; u0(t) is the control input of the system;
[0018] S13. Define the error state limit of the unmanned ship DP system as x∈χ, and the thrust input limit as u0∈U, where the sets χ and U are compact sets containing the origin;
[0019] S14. Define the ocean disturbances such as wind, waves, and surges that the unmanned ship DP system is subjected to as d(t), and the regulated output is z(t). Then the unmanned ship DP control system with propeller failure, ocean disturbances, and constraints is expressed as
[0020]
[0021] in,
[0022] Furthermore, the specific process of step S2 is as follows:
[0023] S21. Based on the full rank decomposition technique, the input matrix B can be decomposed into
[0024] B=B v M
[0025] in, And the rank is l, satisfying l<m;
[0026] S22. Design the integral sliding surface as follows:
[0027]
[0028] Among them, the projection matrix G satisfies GB v =I l ,Right now
[0029] S23. Estimated value of thruster failure factor Update according to the following projection theorem:
[0030]
[0031] Where Γ = -γ 1ι ||α(x)||||M ι ||||u ι ||, Proj{·} is the projection operator;
[0032] S24, design parameters μ0, σ ι , The adaptive estimation law is as follows:
[0033]
[0034]
[0035]
[0036]
[0037] Among them, γ,γ 1ι ,γ 2ι,γ 3ι ,γ 4ι is the adaptive gain determined by the actual situation;
[0038] S25. According to the adaptive parameter estimation law designed in step S24, define the adaptive estimation error: Since the parameters μ0,σ ι ,u si ,d ι The derivatives of are all zero, then the error system can be defined as:
[0039]
[0040] Furthermore, the specific process of step S3 is as follows:
[0041] S31, the integral sliding surface designed according to step S22, taking its derivative, we can get
[0042]
[0043] S32. According to the full rank decomposition technique in step S21, the equation GB=M and the sliding surface definition The equivalent control law can be obtained
[0044]
[0045] Bringing it back to the system state equation, we can get the equivalent sliding mode dynamics:
[0046]
[0047] Furthermore, the specific process of step S4 is as follows:
[0048] S41. According to the element value of the thruster failure factor ρ in the range of [0,1], for the convenience of representation, ρ′=I-ρ can be redefined without changing the range of the element. Discretizing the equivalent sliding mode dynamics, the discrete system can be obtained:
[0049]
[0050] and subject to the mixing constraints
[0051] Fx k +G′u k ≤1,k=1,…,∞
[0052] S42. Design a dual-mode model predictive controller based on the discrete system obtained in step S41
[0053]
[0054] in, is the optimal solution of the MPC optimization problem, K is the feedback matrix, and the Kx term is to reduce the impact of fault uncertainty in the optimization process; define Φ = A d +B d K;
[0055] S43. Decompose the state of the discrete system into
[0056] x k =z k +e k
[0057] Among them, z k is the nominal partial state of the discrete system, e k is a partial state containing fault uncertainty terms; therefore, the discrete system can be decomposed into two parts: the nominal term and the fault uncertainty term:
[0058]
[0059]
[0060] S44. Define variable ι k Parameterize the uncertain tube section of the fault in the form of
[0061] Π k ={e k |Ve k ≤ι k}, where the shape of the tube is determined by selecting the matrix V offline; by including the failure factor The tube parameter ι k As the decision variable of the optimization problem, the fault uncertainty tube can be obtained through online solution; then the dynamic mode 2 can be redefined as
[0062]
[0063] z k+1 =Φz k
[0064] S45, build mode 1 for the nominal state z N The terminal set Z1={z|V f z N ≤1}; Construct a smaller set of terminals for nominal state in mode 2 Construct an invariant set Θ for (z,ι) dynamics k ={ι|||ι|| ∞ ≤τ,ι≥0}.
[0065] Furthermore, the specific process of step S5 is as follows:
[0066] S51. The model predictive control optimization problem based on tube is designed as follows:
[0067]
[0068] stx∈χ,
[0069] u0∈U,
[0070]
[0071]
[0072]
[0073] V f z N ≤1,
[0074]
[0075] S52. In order to prove the feasibility of the optimization problem, the following analysis is performed:
[0076] If the optimization problem is at t k The moment is feasible, that is, there is a feasible solution sequence In t k+1 moment, by iterating over the "tail" sequence It is also a feasible solution to the optimization problem; that is, it proves the feasibility of the optimization problem;
[0077] S53. In order to prove the stability of discrete systems, the prediction cost function is selected As a Lyapunov function, define is the optimal function, we can get
[0078]
[0079] In t k+1 At this moment, the function is a suboptimal function, satisfying You can get
[0080]
[0081] This proves stability.
[0082] Furthermore, the specific process of step S6 is as follows:
[0083] S61. Design an integral sliding mode controller:
[0084]
[0085] in, for Estimate of , μ∈(0,1];
[0086] S62. In order to prove the accessibility and stability of the unmanned ship DP system, a Lyapunov function is constructed, which is in the form of:
[0087]
[0088] in, Derivative V1(t) is calculated and the sliding surface designed in step S22 is substituted into it.
[0089]
[0090] The right side of the above equation is scaled as follows:
[0091]
[0092]
[0093]
[0094]
[0095] Substituting the scaled terms back into the derivative of V1(t), we can get
[0096]
[0097] Combining the adaptive error designed in step S25 and the above formula, the derivative of V can be transformed into
[0098] Substituting κ(t) designed in step S61 into the above formula, we can obtain
[0099]
[0100] Combined with the parameter adaptive estimation law in the design of steps S23 and S24, we can finally get
[0101]
[0102] Right now is a non-increasing function, and when the system is solvable, It is not always zero, which proves that the unmanned ship DP system is asymptotically stable.
[0103] S7. A simulation verification study was conducted on the dynamic model of the unmanned ship DP system with thruster failure using a tube-based model predictive fault-tolerant control scheme, a fault uncertain tube, a state tube, a robust model predictive controller, and an integral sliding mode controller, and compared with conventional means to further verify the effectiveness and superiority.
[0104] Compared with the prior art, the present invention has the following advantages:
[0105] 1. The tube-based model prediction fault-tolerant control method proposed in the present invention can not only ensure the stability of the DP system, but also solve the problem that when a fault occurs, the system state will exceed the range of the terminal set, affecting the stability of the system, compared with the traditional model prediction control strategy. The thruster fault is described as a multiplicative uncertainty process and solved in the tube-based model prediction framework, which not only ensures the robustness of the system in the presence of multiplicative uncertainty, but also reduces conservatism.
[0106] 2. The present invention aims at the multiplicative uncertainty caused by the propeller failure and constructs the fault uncertainty set Π k , nominal system terminal set Z k and the invariant set Θ for the dynamic (z,ι) k , and a new state tube is obtained. And through theoretical analysis, it can be obtained that the occurrence of failure will lead to a reduction in the terminal set.
[0107] 3. The present invention solves the problems of thruster failure and the influence of ocean disturbances through integral sliding mode control based on full rank decomposition and adaptive technology, and handles the multiplicative uncertainty and constraint problems through tube-based model predictive control technology; compared with existing control strategies, the impact of failure on the terminal set is revealed for the first time. BRIEF DESCRIPTION OF THE DRAWINGS
[0108] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative labor.
[0109] Figure 1 The figure is a flow chart of the method of the present invention.
[0110] Figure 2 A state error simulation diagram provided by an embodiment of the present invention.
[0111] Figure 3 This is a control input simulation diagram provided by an embodiment of the present invention.
[0112] Figure 4 This is a simulation diagram of the integral sliding surface provided by an embodiment of the present invention.
[0113] Figure 5 A simulation diagram of the parameter adaptation law provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0114] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme 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 described embodiments 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 creative work should fall within the scope of protection of the present invention.
[0115] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0116] like Figure 1 As shown, the present invention provides a tube-based model predictive control method, characterized in that it includes the following steps:
[0117] S1: Construct the unmanned ship DP system dynamics model based on state and input constraints, thruster fault model and ocean disturbance;
[0118] S2: Design an integral sliding surface according to the state of the unmanned ship DP system dynamics model, and design an adaptive estimation law for the unknown parameters contained in the unmanned ship DP system dynamics model through an adaptive estimation method;
[0119] S3: Based on the definition of the integral sliding mode surface, the equivalent sliding mode dynamics with multiplicative uncertainty caused by the thruster failure factor is established;
[0120] S4: Discretize the equivalent sliding mode dynamics to obtain a discrete system, split the discrete system into a nominal part and an uncertain part, design a tube with fault uncertainty, and use the tube cross-sectional parameters as decision variables to update the size and shape of the tube cross-sectional area online, and design the state invariant set based on the terminal set of the nominal part;
[0121] S5: Design a tube-based model predictive controller and prove the feasibility of the optimization problem in the model predictive controller and the stability of the discrete system;
[0122] S6: Design an integral sliding mode fault-tolerant controller based on the model predictive controller and prove the asymptotic stability of the unmanned ship DP system.
[0123] The specific process of step S1 is as follows:
[0124] S11. Define the position vector of the unmanned ship as ξ(t) = [x p (t)y p (t)φ(t)] T , the velocity vector is Among them, (x p (t)y p (t)) represents the position information, φ(t) represents the heading angle, v(t) and r(t) represent the angular velocity of surge, sway and pitch respectively. Definition of reference state Error Status
[0125] S12. Establish a unified thruster failure model, including three failure modes: thruster failure, interruption and jamming. Its form is
[0126] u F (t)=ρu0(t)+σu s (t)
[0127] Where ρ is a positive definite diagonal matrix, representing the failure factor, and the range of each element is [0,1]; σ is the stuck failure factor, and its value is 0 or 1; u s (t) indicates a stuck fault; u0(t) is the control input of the system;
[0128] S13. Define the error state limit of the unmanned ship DP system as x∈χ, and the thrust input limit as u0∈U, where the sets χ and U are compact sets containing the origin;
[0129] S14. Define the ocean disturbances such as wind, waves, and surges that the unmanned ship DP system is subjected to as d(t), and the regulated output is z(t). Then the unmanned ship DP control system with propeller failure, ocean disturbances, and constraints is expressed as
[0130]
[0131] in,
[0132] The specific process of step S2 is as follows:
[0133] S21. Based on the full rank decomposition technique, the input matrix B can be decomposed into
[0134] B=B v M
[0135] in, And the rank is l, satisfying l<m;
[0136] S22. Design the integral sliding surface as follows:
[0137]
[0138] Among them, the projection matrix G satisfies GB v =I l ,Right now
[0139] S23. Estimated value of thruster failure factor Update according to the following projection theorem:
[0140]
[0141] Where Γ = -γ 1ι ||α(x)||||M ι ||||u ι ||, Proj{·} is the projection operator;
[0142] S24, design parameters μ0, σ ι , The adaptive estimation law is as follows:
[0143]
[0144]
[0145]
[0146]
[0147] Among them, γ,γ 1ι, γ 2ι ,γ 3ι ,γ 4ι is the adaptive gain determined by the actual situation;
[0148] S25. According to the adaptive parameter estimation law designed in step S24, define the adaptive estimation error:
[0149] Since the parameters μ0,σ ι ,u si ,d ι The derivatives of are all zero, then the error system can be defined as:
[0150]
[0151] The specific process of step S3 is as follows:
[0152] S31, the integral sliding surface designed according to step S22, taking its derivative, we can get
[0153]
[0154] S32. According to the full rank decomposition technique in step S21, the equation GB=M and the sliding surface definition The equivalent control law can be obtained
[0155]
[0156] Bringing it back to the system state equation, we can get the equivalent sliding mode dynamics:
[0157]
[0158] The specific process of step S4 is as follows:
[0159] S41. According to the element value of the thruster failure factor ρ in the range of [0,1], for the convenience of representation, ρ′=I-ρ can be redefined without changing the range of the element. The equivalent system after discretization is:
[0160]
[0161] and subject to the mixing constraints
[0162] Fx k +G′u k ≤ 1 ,k=1,…,∞
[0163] S42. Design a dual-mode model predictive controller based on the discrete system obtained in step S41
[0164]
[0165] in, is the optimal solution of the MPC optimization problem, K is the feedback matrix, and the Kx term is to reduce the impact of fault uncertainty in the optimization process. For the convenience of expression, define Φ = A d +B d K;
[0166] S43. Decompose the state of the discrete system into
[0167] x k =z k +e k
[0168] Among them, z k is the nominal partial state of the discrete system, e k is a partial state containing fault uncertainty terms; therefore, the discrete system can be decomposed into two parts: the nominal term and the fault uncertainty term:
[0169]
[0170]
[0171] S44. Define variable ι k Parameterize the uncertain tube section of the fault in the form of Π k ={e k |Ve k ≤ι k}, where the shape of the tube is determined by selecting the matrix V offline; by including the failure factor The tube parameter ι k As the decision variable of the optimization problem, the fault uncertainty tube can be obtained through online solution; then the dynamic mode 2 can be redefined as
[0172]
[0173] z k+1 =Φz k
[0174] S45, build mode 1 for the nominal state z N The terminal set Z1={z|V f z N ≤1}; Construct a smaller set of terminals for nominal state in mode 2 Construct an invariant set Θ for (z,ι) dynamics k ={ι|||ι|| ∞ ≤τ,ι≥0}.
[0175] The specific process of step S5 is as follows:
[0176] S51. The model predictive control optimization problem based on tube is designed as follows:
[0177]
[0178] stx∈χ,
[0179] u0∈U,
[0180]
[0181]
[0182]
[0183] V f z N ≤1,
[0184]
[0185] S52. In order to prove the feasibility of the optimization problem, the following analysis is performed:
[0186] If the optimization problem is at t k The moment is feasible, that is, there is a feasible solution sequence In t k+1 moment, by iterating over the "tail" sequence It is also a feasible solution to the optimization problem; that is, it proves the feasibility of the optimization problem;
[0187] S53. In order to prove the stability of discrete systems, the prediction cost function is selected As a Lyapunov function, define is the optimal function, we can get
[0188]
[0189] In t k+1 At this moment, the function is a suboptimal function, satisfying You can get
[0190]
[0191] This proves stability.
[0192] The specific process of step S6 is as follows:
[0193] S61. Design an integral sliding mode controller:
[0194]
[0195] in, for Estimate of , μ∈(0,1];
[0196] S62. In order to prove the accessibility and stability of the unmanned ship DP system, a Lyapunov function is constructed, which is in the form of:
[0197]
[0198] in, Derivative V1(t) is calculated and the sliding surface designed in step S22 is substituted into it.
[0199]
[0200] The right side of the above equation is scaled as follows:
[0201]
[0202]
[0203]
[0204]
[0205] Substituting the scaled terms back into the derivative of V1(t), we can get
[0206]
[0207] Combining the adaptive error designed in step S25 and the above formula, the derivative of V can be transformed into
[0208] Substituting κ(t) designed in step S61 into the above formula, we can obtain
[0209]
[0210] Combined with the parameter adaptive estimation law in the design of steps S23 and S24, we can finally get
[0211]
[0212] Right now is a non-increasing function, and when the system is solvable, It is not always zero, which proves that the unmanned ship DP system is asymptotically stable.
[0213] In addition to steps 1-6, it also includes:
[0214] S7. Conduct simulation verification studies on the dynamic model of the unmanned ship dynamic positioning system with thruster failures and constraints, tube-based model prediction and integral sliding mode collaborative controller, integral sliding surface, adaptive estimation law and fault uncertain tube, and compare them with conventional means to further verify the effectiveness and superiority.
[0215] In order to verify the effectiveness of the tube-based model predictive control method provided in this embodiment, a simulation experiment is carried out using MATLAB, and a detailed description is given.
[0216] Taking into account input and state constraints, thruster failures, ocean disturbances and multiplicative uncertainty factors, model prediction and integral sliding mode control techniques were adopted, and a tube-based model predictive controller was designed. It can keep the system state within the tube, is robust to thruster failures and ocean disturbances, reduces conservatism, and ensures tracking performance.
[0217] Specifically, in this embodiment, a model of a typical floating production vessel scaled at a ratio of 70:1 is used. In the simulation, the initial conditions of the initial state and parameters of the vessel are set as:
[0218] x0=[0.20 -0.10 0.20 0.10 -0.30 0.2] T ,
[0219] The adaptive gain is selected as γ = γ 1i =γ 2i =γ 3i =γ 4i = 0.1; the constraints of system state and reasoning input are set to The selected ocean disturbance form is d(t) = [0.3sin(0.2t)-0.15 0.35 0.2cos(0.2t)-0.15cos(0.15t)0.15] T .
[0220] In the simulation, the sampling time is set to 0.1s, the weight matrix Q = I, the weight coefficient R = 0.1; the prediction step size N = 6, N m =10; the time range of the ocean disturbance is set at t=15s to t=20s; the failure starts from t=15s, the second thruster is set to fail, and the sixth thruster is set to interrupt at 0.2+0.1sin(t).
[0221] Based on the above parameters, the tube-based unmanned ship model prediction fault-tolerant control method proposed in the present invention is simulated and verified. Figure 3-5 As shown. Among them, Figure 2 It is shown as the error state of the system. It can be seen from the figure that it finally converges to 0, that is, the unmanned ship can achieve DP and the system is stable; Figure 3 The controller of the display system shows that the thrust of the six thrusters eventually tends to 0, which means that the controller is effective; Figure 4 It shows that the integral sliding mode curve eventually tends to 0, and the vibration can be suppressed to almost Tsuen Wan hours; Figure 5 It shows that the parameter adaptive estimation gradually tends to a stable constant. So far, the digital simulation of the algorithm has been completed and its effectiveness has been verified.
[0222] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.
[0223] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0224] 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned 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 tube-based unmanned ship model predictive control method, characterized in that: The steps include: S1: Construct the unmanned ship DP system dynamics model based on state and input constraints, thruster fault model and ocean disturbance; S2: Design an integral sliding surface according to the state of the unmanned ship DP system dynamics model, and design an adaptive estimation law for the unknown parameters contained in the unmanned ship DP system dynamics model through an adaptive estimation method; The specific process of step S2 is as follows: S21. Based on the full rank decomposition technique, the input matrix B is decomposed into B=B v M in, And the rank is l, satisfying l <m; S22. Design the integral sliding surface as follows: Among them, the projection matrix G satisfies GB v =I l ,Right now S23. Estimated value of thruster failure factor Update according to the following projection theorem: Where Γ = -γ 1ι ||α(x)||||M ι ||||u ι ||, Proj{·} is the projection operator; S24, design parameters μ0, σ ι , The adaptive estimation law is as follows: Among them, γ,γ 1ι ,γ 2ι ,γ 3ι ,γ 4ι is the adaptive gain determined by the actual situation; S25. Define the adaptive estimation error: Since the parameters μ0,σ ι ,u si ,d ι The derivatives of are all zero, then the error system is defined as: S3: Based on the definition of the integral sliding mode surface, the equivalent sliding mode dynamics with multiplicative uncertainty caused by the thruster failure factor is established; S4: Discretize the equivalent sliding mode dynamics to obtain a discrete system, split the discrete system into a nominal part and an uncertain part, design a tube with fault uncertainty, and use the tube section parameters as decision variables to update the size and shape of the tube section online, and design the state invariant set based on the terminal set of the nominal part; S5: Design a tube-based model predictive controller and prove the feasibility of the optimization problem in the model predictive controller and the stability of the discrete system; S6: Design an integral sliding mode fault-tolerant controller based on the model predictive controller and prove the asymptotic stability of the unmanned ship DP system.
2. The tube-based model prediction fault-tolerant control method according to claim 1, characterized in that: The specific process of step S1 is as follows: S11: Define the reference state of the unmanned ship as x ref , the error state is x(t); S12. Establish a unified thruster failure model, including three failure modes: thruster failure, interruption and jamming. Its form is u F (t)=ρu0(t)+σu s (t) Where ρ is a positive definite diagonal matrix, representing the failure factor, and the range of each element is [0,1]; σ is the stuck failure factor, and its value is 0 or 1; u s (t) indicates a stuck fault; u0(t) is the control input of the system; S13. Define the error state limit of the unmanned ship DP system as x∈χ, and the thrust input limit as u0∈U, where the sets χ and U are compact sets containing the origin; S14. Define the ocean disturbances of wind, waves and surges to the unmanned ship DP system as d(t), and the regulated output z(t). Then the unmanned ship DP control system with propeller failure, ocean disturbance and constraints is expressed as in, 3. The tube-based model prediction fault-tolerant control method according to claim 2, characterized in that: The specific process of step S3 is as follows: S31, according to the integral sliding surface designed in step S22, take its derivative to obtain S32. According to the full rank decomposition technique in step S21, the equation GB=M and the sliding surface definition The equivalent control law can be obtained Bringing it back to the system state equation, we can get the equivalent sliding mode dynamics:
4. The tube-based model prediction fault-tolerant control method according to claim 3, characterized in that: The specific process of step S4 is as follows: S41. According to the element value of the thruster failure factor ρ in the range of [0,1], ρ′=I-ρ is redefined without changing the range of the element, and the equivalent sliding mode dynamics is discretized to obtain the discrete system: and subject to the mixing constraints Fx k +G′u k ≤ 1 ,k=1,…,∞ S42. Design a dual-mode model predictive controller based on the discrete system obtained in step S41 in, is the optimal solution of the MPC optimization problem, K is the feedback matrix, and the Kx term is to reduce the impact of fault uncertainty in the optimization process; and define Φ = A d +B d K; S43. Decompose the state of the unmanned ship DP system into x k =z k +e k Among them, z k is the nominal partial state, e k It is a partial state containing the uncertainty item of fault; S44. Define variable ι k Parameterized fault uncertainty tube cross section, form Π k ={e k |Ve k ≤ι k }, where the shape of the tube is determined by selecting the matrix V offline; by including the failure factor The tube parameter ι k As the decision variable of the optimization problem, the fault uncertainty tube is obtained through online solution; then the dynamic mode 2 is redefined as z k+1 =Φz k S45, build mode 1 for the nominal state z N The terminal set Z1={z|V f z N ≤1}; Construct a smaller set of terminals for nominal state in mode 2 Construct an invariant set Θ for (z,ι) dynamics k ={ι|||ι|| ∞ ≤τ,ι≥0}.
5. The tube-based model prediction fault-tolerant control method according to claim 4, characterized in that: The specific process of step S5 is as follows: S51. The model predictive control optimization problem based on tube is designed as follows: P: stx∈χ, u0∈U, In f with N ≤ 1 , S52. In order to prove the feasibility of the optimization problem, the following analysis is performed: If the optimization problem is in t k The moment is feasible, that is, there is a feasible solution sequence In t k+1 moment, by iterating over the "tail" sequence It is also a feasible solution to the optimization problem; that is, it proves the feasibility of the optimization problem; S53. In order to prove the stability of discrete systems, the prediction cost function is selected As a Lyapunov function, define is the optimal function, we can get In t k+1 At this moment, the function is a suboptimal function, satisfying You can get This proves stability.
6. The tube-based model predictive fault-tolerant control method according to claim 1 or 2, characterized in that: The specific process of step S6 is as follows: S61. Design an integral sliding mode controller: in, for Estimate of , μ∈(0,1]; S62. In order to prove the accessibility and stability of the unmanned ship DP system, a Lyapunov function is constructed, which is in the form of: in, Derivative V1(t) is calculated and the sliding surface designed in step S22 is substituted into it. After scaling the terms on the right side of the above equation, and combining the adaptive error designed in step S25 and the above equation, the derivative of V can be transformed into Substituting κ(t) designed in step S61 into the above formula, we obtain Combined with the parameter adaptive estimation law in the design of steps S23 and S24, we finally get Right now is a non-increasing function, and when the system is solvable, It is not always zero, which proves that the unmanned ship DP system is asymptotically stable.
