Unmanned ship robust dynamic output feedback prediction control method

Abstract

The invention relates to a robust dynamic output feedback prediction control method for an unmanned ship. The method comprises the following steps: establishing an interval type-2 fuzzy model of the unmanned ship; establishing an optimization problem of on-line state feedback robust predictive control, and solving the optimization problem to obtain an on-line state feedback predictive control algorithm; based on the interval type-2 fuzzy model, establishing an interval type-2 observer of the unmanned ship; and performing dynamic robust model prediction on the unmanned ship through the online state feedback prediction control algorithm and the dynamic output feedback prediction control algorithm so as to realize stabilization control of the unmanned ship. According to the invention, the accuracy in anti-interference stabilization control of the unmanned ship can be improved, the system can be stabilized efficiently and robustly, and the safety of the unmanned ship in a complex environment is ensured.

Description

Technical Field

[0001] The invention relates to the technical field of unmanned boat control, and in particular to an unmanned boat robust dynamic output feedback prediction control method. Background Art

[0002] Unmanned boats are small intelligent marine transport platforms with remote control or autonomous navigation capabilities. They integrate many modern high-tech technologies. They have the ability to perform tasks around the clock, especially in harsh marine environments, replacing humans to perform dangerous, heavy and time-consuming tasks. They are widely used in many fields and have broad prospects. With the rapid development of unmanned driving technology, unmanned boats have attracted more and more attention due to their advantages such as autonomy, flexibility, small size and low cost, especially in reducing casualties and performing high-risk tasks. Therefore, it is of great practical significance to design a feedback control strategy for unmanned boats.

[0003] However, the current feedback control strategy ignores the restoring force matrix, and the stabilization control of the unmanned boat does not introduce the membership function information, which will lead to poor control strategy effects in practical applications, and fail to effectively respond to environmental changes and maintain system stability, thus affecting the stabilization performance of the system. Summary of the invention

[0004] The present invention provides an unmanned boat robust dynamic output feedback predictive control method to solve the defects of the prior art.

[0005] The present invention provides a method for robust dynamic output feedback predictive control of an unmanned boat, comprising:

[0006] S1: Establish an interval type-2 fuzzy model of the unmanned boat;

[0007] S2: Establish an optimization problem of online state feedback robust predictive control, and obtain an online state feedback predictive control algorithm by solving the optimization problem;

[0008] S3: Based on the interval type-2 fuzzy model, an interval type-2 observer of the unmanned boat is established;

[0009] S4: Through the online state feedback predictive control algorithm and the dynamic output feedback predictive control algorithm, a dynamic robust model prediction is performed on the unmanned boat to achieve stabilization control of the unmanned boat.

[0010] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S1 further comprises:

[0011] S11: Based on the surface motion characteristics of the unmanned boat, the state space model of the unmanned boat is established;

[0012] S12: Based on the state space model, a basic type-2 fuzzy model is established;

[0013] S13: Obtain an interval type-2 fuzzy model by discretizing the basic type-2 fuzzy model.

[0014] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S11 further comprises:

[0015] S111: Select the three-degree-of-freedom unmanned boat dynamics model as the basic model;

[0016] S112: Based on the stabilization control problem of the unmanned boat, the centripetal force and the Coriolis force matrix of the basic model are ignored to obtain a simplified model;

[0017] S113: Based on the correspondence between the state and position of the unmanned boat and the correspondence between the state and the navigation speed, the simplified model is converted into a state space model in the form of a state equation.

[0018] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S12 further comprises:

[0019] S121: Based on the range of the heading angle of the unmanned boat, define the second type of fuzzy rules of the unmanned boat;

[0020] S122: define upper and lower bounds of membership weight functions to obtain a type II membership weight function;

[0021] S123: Based on the type-2 fuzzy rules and the type-2 membership weight function, a basic type-2 fuzzy model is obtained.

[0022] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S2 further comprises:

[0023] S21: Establish the optimization problem of online state feedback robust predictive control;

[0024] S22: Introducing a robust positive invariant set into the optimization problem to obtain an update problem;

[0025] S23: A method for solving the update problem based on the robust positive invariant set and the membership function information to obtain an online state feedback predictive control algorithm.

[0026] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S21 further comprises:

[0027] S211: Characterize the input control torque of the unmanned boat and the state constraint of the unmanned boat to obtain the input control torque and the state constraint;

[0028] S212: Based on the non-parallel allocation compensation method, an interval type-II observer for the unmanned boat is established;

[0029] S213: According to the input control torque, the state constraint, and the interval type-II observer, an infinite time domain unmanned boat stabilization control optimization problem is established to obtain an optimization problem.

[0030] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S23 further includes:

[0031] S231: Convert Lyapunov attenuation and robust performance constraints into first matrix inequality and second matrix inequality respectively;

[0032] S232: based on the second matrix inequality combined with the robust positive invariant set, according to the input control torque and the state constraint, the control input constraint and the state constraint are respectively converted into a third matrix inequality and a fourth matrix inequality;

[0033] S233: Based on the membership function vertex envelope method, membership function information is introduced into the first matrix inequality, the third matrix inequality, and the fourth matrix inequality to obtain a membership matrix inequality;

[0034] S234: Solving the update problem based on the membership matrix inequality to obtain an online state feedback predictive control algorithm.

[0035] According to a method for robust dynamic output feedback predictive control of an unmanned boat provided by the present invention, the online state feedback predictive control algorithm in step S234 specifically includes:

[0036] S2341: Initialize system status;

[0037] S2342: Solve the update problem to obtain a state feedback sub-gain;

[0038] S2343: combining the state feedback sub-gain with the membership function to obtain a local control law;

[0039] S2344: Input the control signal corresponding to the local control law into the control system, and perform update control according to the time step.

[0040] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S3 further comprises:

[0041] S31: Based on the system output when the state of the unmanned boat system is unmeasurable, an interval type II Lumberg observer is established;

[0042] S32: constructing an error system based on the interval type-2 fuzzy model and the interval type-2 Lumberg observer;

[0043] S33: According to the error evolution trajectory corresponding to the error system, the system state is expanded to obtain an interval type-II observer based on the observation error type-II augmented system.

[0044] According to the unmanned boat robust dynamic output feedback predictive control method provided by the present invention, step S4 further comprises:

[0045] S41: establishing a type-II feedback control rate according to the estimated state of the system, substituting the type-II feedback control rate into the interval type-II observer, and obtaining a basic dynamic output feedback system;

[0046] S42: Introducing the robust positive invariant set and the membership function information into the basic dynamic output feedback system to obtain an updated dynamic output feedback system;

[0047] S43: Solving the updated dynamic output feedback system through an online state feedback predictive control algorithm and a corresponding dynamic output feedback predictive control algorithm to perform robust dynamic output feedback predictive control on the unmanned boat.

[0048] The present invention provides a method for robust dynamic output feedback predictive control of an unmanned boat. When establishing an interval type-II fuzzy model of the unmanned boat, the surface motion characteristics of the unmanned boat are fully considered. By including a restoring force matrix, the model of the present invention is more accurate and can better reflect the actual motion characteristics of the unmanned boat, thereby improving the effectiveness and stability of the control strategy. The present invention introduces membership function information into the optimization problem and the control algorithm. By introducing the membership function information, the present invention can more accurately describe the state of the unmanned boat system and the uncertainty of the control input, thereby improving the performance of the stabilization control. At the same time, the present invention also ensures that the control algorithm can be designed through an efficient algorithm. Adjustments are made within the specified time to avoid problems such as response lag or system loss of control; the present invention fully considers various constraints of the unmanned boat system in the control strategy, including physical constraints such as linear velocity, angular velocity, thrust, and rudder deviation. By considering these constraints, the present invention can design a control algorithm that is more in line with practical applications and avoid safety problems caused by ignoring constraints; the present invention adopts an online state feedback robust predictive control algorithm, which can be dynamically adjusted according to the real-time state of the unmanned boat, thereby improving the robustness and adaptability of the system. In addition, by introducing robust positive invariant sets and membership function information, the present invention further enhances the stability and performance of the system.

[0049] The present invention provides new ideas and methods for the development of unmanned boat control technology. By combining fuzzy control, predictive control and other technologies, it provides strong support for the realization of autonomous navigation, intelligent obstacle avoidance, collaborative operation and other aspects of unmanned boats. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the present invention or 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 work.

[0051] Figure 1 A schematic flow chart of a robust dynamic output feedback predictive control method for an unmanned boat provided by the present invention;

[0052] Figure 2 A schematic diagram of the flow of the interval type-2 fuzzy model acquisition method provided by the present invention;

[0053] Figure 3 A schematic diagram of a flow chart of a method for acquiring an online state feedback predictive control algorithm provided by the present invention;

[0054] Figure 4 A schematic diagram of the flow of the interval type-2 observer acquisition method provided by the present invention;

[0055] Figure 5 A schematic flow chart of a method for predicting a dynamic output feedback robust model of an unmanned boat provided by the present invention;

[0056] Figure 6 A schematic diagram of the actual state of the unmanned boat under the state feedback predictive control provided by the present invention;

[0057] Figure 7 A schematic diagram of the actual state of the unmanned boat under the output feedback predictive control provided by the present invention;

[0058] Figure 8 A schematic diagram of the observation state of the unmanned boat under the output feedback predictive control provided by the present invention;

[0059] Fig. 9 This is a schematic diagram of the control input of the output feedback predictive control unmanned boat provided by the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0061] In order to better understand the embodiments of the present invention, the research background of the present invention is first explained in detail below.

[0062] The motion characteristics of unmanned boats are significantly challenging. First, the mathematical model of the motion of unmanned boats is the key to designing and analyzing controllers. Due to its nonlinear characteristics, the modeling process is complex and costly. In addition, there are a lot of uncertainties in the motion model of unmanned boats, including uncertainty in model parameters, unmodeled dynamics, and wind, wave and current disturbances in the marine environment. The control system of unmanned boats is usually under-actuated. The control input of the propulsion system is less than its degrees of freedom. It is a typical under-actuated system, which makes it impossible to solve its motion model by traditional feedback linearization methods. Brockett's theorem shows that it is impossible to achieve fixed-point regulation through a time-invariant, smooth state feedback controller. Due to the limitation of the drive capacity, the motion model of the unmanned boat must consider physical constraints such as speed, angular velocity, thrust and rudder deviation. Ignoring these constraints may lead to a decrease in control performance or even system instability. In addition, in practical applications, the position information of the unmanned boat can be obtained through positioning systems such as GPS, but the speed information cannot be directly measured. Although accelerometers can provide acceleration information, they cannot directly obtain speed. Although Doppler logs can measure speed, their high cost is not suitable for large-scale applications, especially low-cost small unmanned boats. Therefore, designing an effective speed observer and feedback control strategy has important practical significance, which can significantly reduce the implementation cost of the control algorithm. It is particularly important to design a control algorithm that comprehensively considers the characteristics of the unmanned boat.

[0063] For the current unmanned boat control method, the existing unmanned boat interval type-II modeling technology ignores the restoring force matrix. However, omitting the restoring force will lead to poor control strategy in practical applications, and it will not be able to effectively respond to environmental changes and maintain system stability.

[0064] Secondly, the existing unmanned boat stabilization control does not introduce membership function information, and the membership function contains key information of the unmanned boat. Ignoring this will affect the stabilization performance of the system. While introducing the membership function, it is also necessary to ensure the efficiency of the algorithm. Otherwise, if the control algorithm fails to make adjustments within the specified time, it may cause response lag or system out of control.

[0065] In addition, current technologies do not take system constraints into account. The mathematical model of the unmanned boat movement contains physical constraints such as linear velocity, angular velocity, thrust, and rudder deviation, which are related to the driving ability of the unmanned boat and the system state limitations. Therefore, ignoring these constraints will lead to the inability to ensure the safety of the unmanned boat, making the control algorithm unable to play a role in practical applications.

[0066] Finally, most of the existing unmanned boat control systems are based on state feedback, but since some states of the unmanned boat cannot be directly measured in a low-cost way, this brings challenges to the design and application of the control system.

[0067] The embodiments of the present invention are described below with reference to the accompanying drawings.

[0068] like Figure 1 As shown, the present invention provides a method for robust dynamic output feedback predictive control of an unmanned boat, comprising:

[0069] S1: Establish an interval type-2 fuzzy model of the unmanned boat.

[0070] The purpose of step S1 of the present invention is to establish a mathematical model that can handle uncertainty and fuzziness for the dynamic behavior of the unmanned boat. Compared with the traditional fuzzy model, the interval type-2 fuzzy model established in step S1 can better handle the uncertainty in input and output, thereby improving the robustness of the control system.

[0071] like Figure 2 As shown, step S1 further includes:

[0072] S11: Based on the surface motion characteristics of the unmanned boat, a state space model of the unmanned boat is established.

[0073] Wherein, step S11 further comprises:

[0074] S111: Select the three-degree-of-freedom unmanned boat dynamics model as the basic model.

[0075] Furthermore, the present invention considers the dynamic model of a three-degree-of-freedom USV, and the specific expression is:

[0076]

[0077] η=[η x ,η t ,η ψ ]∈R 3 is the position vector of USV, (η x ,η y ) represents the position in the geodetic coordinate system, η ψ is the heading angle in the geodetic coordinate system, v = [v x ,v y ,v ψ ]∈R 3 is the navigation speed vector of the USV, where v x Refers to the longitudinal velocity, v y Refers to the sway velocity, v ψ refers to the bow speed, τ refers to the desired control input, S∈R 3×3 is a positive definite symmetric inertia matrix, C(v)∈R3×3 Represents the centripetal force and Coriolis force matrix, D(v)∈R 3×3 represents the damping matrix, G is the restoring force caused by gravity, buoyancy and ocean currents, ω refers to the unknown disturbance and model uncertainty, and γ(η) refers to the non-singular transformation matrix from the hull coordinate system to the earth coordinate system.

[0078] S112: Based on the stabilization control problem of the unmanned boat, the centripetal force and the Coriolis force matrix of the basic model are ignored to obtain a simplified model.

[0079] In the stabilization control problem, in order to simplify the model and reduce the complexity of the control design, the present invention chooses to ignore the centripetal force and the Coriolis force matrix. These forces may be more significant during high-speed motion or large turns, but in the stabilization control problem, in order to grasp the main contradictions, they can be ignored to appropriately simplify the model.

[0080] Specifically, the present invention considers the stabilization control problem of unmanned boats. The control goal is to stabilize the ship to a set equilibrium point and maintain the desired direction, which is actually manifested as positioning control, automatic berthing, etc. For under-actuated ships, since the system is subject to a second-order non-holonomic constraint in which the acceleration is not integrable, there is no smooth time-invariant feedback control rate to make the system asymptotically stable.

[0081] For the centripetal force and Coriolis force matrix, on the one hand, when the unmanned boat is sailing at a low speed on the water surface, the resistance and speed can usually be approximated to be a linear relationship. At this time, the hydrodynamic resistance term in the dynamic model can be constructed in a linear form. On the other hand, when modeling, considering that the unmanned boat mainly moves in the horizontal plane, the vertical, roll and pitch motion of the unmanned boat are ignored, thereby simplifying the six-degree-of-freedom hydrodynamic equation to a three-degree-of-freedom hydrodynamic equation in the horizontal plane. In this case, the influence of the centripetal force and Coriolis force is relatively small, and the influence of these forces can be considered to be simplified or ignored. Therefore, the centripetal force and Coriolis force matrix are omitted, and the unmanned boat motion model on the water surface becomes:

[0082]

[0083] S113: Based on the correspondence between the state and position of the unmanned boat and the correspondence between the state and the navigation speed, the simplified model is converted into a state space model in the form of a state equation.

[0084] Specifically, the present invention defines the corresponding relationship between the state and the position and navigation speed of the unmanned boat in step S113, that is, x1=η, x2=v, u=τ, then the general USV dynamic system is expressed as:

[0085]

[0086] Expressed in the form of state equation:

[0087]

[0088] In the formula,

[0089] In step S113, the present invention converts the simplified unmanned boat dynamics model into a state space model in the form of a state equation. The state space model is a mathematical tool that describes the dynamic behavior of a system. It uses state variables to describe the internal state of the system, and uses state equations to describe how these state variables change over time. Step S113 is for the subsequent use of modern control theories and methods to design a controller.

[0090] S12: Based on the state space model, a basic type-2 fuzzy model is established.

[0091] Wherein, step S12 further comprises:

[0092] S121: Based on the variation range of the heading angle of the unmanned boat, define the type II fuzzy rules of the unmanned boat.

[0093] The present invention considers the heading angle in Internal changes, resulting from Define the interval type II fuzzy rules of the unmanned boat, specifically:

[0094] Rule i: If sin(η ψ ) is M i1 , and cos(η ψ ) is M i1 ,So:

[0095]

[0096]

[0097] In step S121, the present invention defines a type-II fuzzy rule according to the range of variation of the heading angle of the unmanned boat. Compared with the traditional type-I fuzzy rule, the type-II fuzzy rule can better handle ambiguity and uncertainty. The provided rules describe the possible behaviors and responses of the unmanned boat under different heading angle states.

[0098] S122: Define upper and lower bounds of membership weight functions to obtain a type II membership weight function.

[0099] In step S122, the upper and lower bounds of the membership weight function are first defined, and then the average centroid method is used to defuzzify, and the following is obtained:

[0100] The activation strength of the i-th fuzzy rule can be expressed as:

[0101]

[0102] The membership function of the i-th fuzzy rule is:

[0103]

[0104] in, is a predefined nonlinear weight function that satisfies

[0105] In step S122, the present invention defines upper and lower bound membership weight functions to describe the uncertainty of membership in a type-2 fuzzy set. Through these weight functions, the present invention can more accurately describe the uncertainty of the state of the unmanned boat.

[0106] S123: Based on the type-2 fuzzy rules and the type-2 membership weight function, a basic type-2 fuzzy model is obtained.

[0107] Furthermore, by combining the results of the above steps S121 and S122, a type II fuzzy system can be obtained, and the specific expression is:

[0108]

[0109] The obtained model can describe the fuzzy behavior and uncertainty of the unmanned boat in different states.

[0110] S13: Obtain an interval type-2 fuzzy model by discretizing the basic type-2 fuzzy model.

[0111] Furthermore, the application of discrete control in modern control systems is inevitable. It provides an effective way to achieve efficient, reliable and flexible control strategies. The zero-order holder is superior to the Euler method in terms of accuracy and stability. If the sampling time is T, according to the parameter matrix of the type II model system, the restoring force matrix G is often a singular matrix, which leads to A being not full of rank. The Euler method is used instead. The specific expression is:

[0112] In actual control systems, due to the limitations of computer processing capabilities and real-time requirements, continuous-time models often need to be discretized. Therefore, in step S13, the researchers discretized the basic type-2 fuzzy model and obtained an interval type-2 fuzzy model. The interval type-2 fuzzy model allows the membership function to vary within a certain interval, so it has a stronger ability to handle uncertainty and ambiguity.

[0113] S2: Establish the optimization problem of online state feedback robust predictive control, and obtain the online state feedback predictive control algorithm by solving the optimization problem.

[0114] like Figure 3As shown, step S2 further includes:

[0115] S21: Establish the optimization problem of online state feedback robust predictive control.

[0116] Wherein, step S21 further includes:

[0117] S211: Characterize the input control torque of the unmanned boat and the state constraint of the unmanned boat to obtain the input control torque and the state constraint.

[0118] In step S211, the present invention first formally characterizes the control input torque and state constraints of the unmanned boat, assuming that the control input torque and state constraints of the unmanned boat are as follows:

[0119]

[0120] Among them, u max is the upper bound of the control input, [*] α is the αth element in *, M∈R 1×n , ξ represents the number of restrictions on the system state.

[0121] S212: Based on the non-parallel allocation compensation method, an interval type-II observer for the unmanned boat is established.

[0122] Secondly, in step S212, the present invention designs a non-parallel distribution compensation method to design an interval type II fuzzy controller for the unmanned boat. The non-parallel distribution compensation method is used to design an interval type II fuzzy controller for the unmanned boat as follows:

[0123] u=K n x;

[0124]

[0125] in, is a predefined nonlinear weight function that satisfies Compared to Take it as a fixed value, As a nonlinear function, the conservatism of the controller design can be reduced.

[0126] S213: According to the input control torque, the state constraint, and the interval type-II observer, an infinite time domain unmanned boat stabilization control optimization problem is established to obtain an optimization problem.

[0127] In step S213, the infinite time domain unmanned boat stabilization control optimization problem is established by combining the results obtained in step S211 and step S212. The specific expression is:

[0128]

[0129] in, is the objective function, Q>0, R>0 is the known weight matrix, and *(k+c|k) represents the value of k+c at the future moment predicted using the current information.

[0130] S22: Introduce a robust positive invariant set into the optimization problem to obtain an update problem.

[0131] In step S22, a robust positive invariant set is introduced to transform the min / max problem of robust predictive control optimization in infinite time domain in the above step S213 into a min problem of minimizing the upper bound of the performance index. The specific process is as follows.

[0132] First, define the quadratic Lyapunov function as:

[0133] V(k)=x T (k|k)P x x(k|k);

[0134] Where P x >0, assuming that the difference between the predicted Lyapunov function at k+c and k+c+1 satisfies:

[0135] V(k+c+1|k)-V(k+c|k)+x(k+c|k) T Qx(k+c|k)+u(k+c|k) T Ru(k+c|k)≤0;

[0136] Add the above formula from c = 0 to c = ∞, lim c→∞ x(k+c|k)=0, when lim c→∞ When V(x(k+c|k))=0, we can get Considering V(k|k)≤γ, we introduce H ∞ Performance index: ∥y(k)∥2≤γ c ∥ω∥2, the optimization problem can be updated as:

[0137]

[0138] That is the min problem of minimizing the upper bound of the performance index mentioned above.

[0139] S23: A solution method for the update problem based on the robust positive invariant set and the membership function information is used to obtain an online state feedback predictive control algorithm.

[0140] Wherein, step S23 further comprises:

[0141] S231: Convert the Lyapunov attenuation and robust performance constraints into the first matrix inequality and the second matrix inequality respectively.

[0142] In step S231, the Lyapunov attenuation and robust performance constraints are first converted into matrix inequalities. The Lyapunov attenuation under the robust constraint can be guaranteed by the following formula:

[0143]

[0144] In the formula Matrix pre- and post-multiplication diag{N x γ,I}, define Y n =K n N x , And using Schur complement we get:

[0145]

[0146] S232: Based on the second matrix inequality combined with the robust positive invariant set, according to the input control torque and the state constraint, the control input constraint and the state constraint are respectively converted into a third matrix inequality and a fourth matrix inequality.

[0147] Furthermore, in step S232, it is necessary to combine the robust positive invariant set to transform the control input constraints and the state constraints into matrix inequalities. The specific steps are as follows.

[0148] For terminal restrictions, at the current moment, if x(k|k) T P x If x(k|k)≤γ, then the terminal moment must satisfy V(k+N|k)≤γ, which can be converted into a linear expression using Schur complement.

[0149] For the control input limits, we have:

[0150]

[0151] The above formula is valid if:

[0152]

[0153] Further, multiply N before and after x , we can get:

[0154]

[0155] For status restrictions, there are:

[0156] max{[Mx(k+c|k)] β |x(k+c|k)∈Ω x}≤1;

[0157] Using the Lagrange multiplier method, it can be ensured by the following formula:

[0158]

[0159] Use Schur complement and multiply N before and after x , we can get Ω≤0.

[0160] S233: Based on the membership function vertex envelope method, membership function information is introduced into the first matrix inequality, the third matrix inequality, and the fourth matrix inequality to obtain membership matrix inequalities.

[0161] In the unified space of membership functions, the vertex envelope method is adopted and the following algorithm is used to calculate the envelope vertices of the product of type-two membership functions in the membership function space, thereby introducing the information of type-two membership functions into the Lyapunov function decreasing inequality, state restriction, and sufficient guarantee inequality of input constraints, thereby reducing the conservatism of the solution.

[0162] The specific interval type II membership function space vertex envelope algorithm is: Calculate f i The maximum and minimum values ​​of (x), f imin =min{f i (z)},f imax =max{f i (z)}; calculate the maximum and minimum values ​​of the new membership defined as follows: Create variable β svj :β sv0 =α s0 , Final Order Subsequently, the coordinates of the convex envelope vertices are obtained through base coordinate transformation and modulus operation, and the envelope points are used to solve the above matrix inequality.

[0163] S234: Solving the update problem based on the membership matrix inequality to obtain an online state feedback predictive control algorithm.

[0164] According to the results in steps S21 to S22, the present invention obtains the following unmanned boat online state feedback predictive control algorithm.

[0165] The online state feedback predictive control algorithm in step S234 specifically includes:

[0166] S2341: Initialize system status.

[0167] S2342: Solve the update problem to obtain a state feedback sub-gain.

[0168] S2343: Combine the state feedback sub-gain with the membership function to obtain a local control law.

[0169] S2344: Input the control signal corresponding to the local control law into the control system, and perform update control according to the time step.

[0170] The specific online state feedback predictive control algorithm is as follows: initialize the system state online, solve the optimization problem in step S21 (the update problem in step S22), and obtain the state feedback sub-gain K i =Y i (N x ) -1 , and then combine these sub-gains with the membership functions (MFs) to obtain the local control law u(k). If it is not satisfied, u(k) = 0, and the control signal u(k) is input into the system, then k is updated to k+1 and then the control signal u(k+1) is returned and input into the system, and it is continuously updated.

[0171] The premise for the optimization problem established in step S2 is that the state of the unmanned boat system can be measured. However, in practical applications, the position information of the unmanned boat can be obtained through cheap global navigation and positioning systems such as GPS and Beidou positioning systems, but its speed information cannot be directly measured through the navigation and positioning system. The accelerometer can only measure acceleration information, not speed information. Although the Doppler log can directly measure the speed information of the unmanned boat, it is expensive and not suitable for large-scale unmanned boat cluster applications, especially small low-cost unmanned boats. Therefore, it is of practical significance to study the speed observer and its output feedback control problem, which can significantly reduce the implementation cost of the control algorithm.

[0172] S3: Based on the interval type-2 fuzzy model, an interval type-2 observer of the unmanned boat is established.

[0173] like Figure 4 As shown, step S3 further includes:

[0174] S31: Based on the system output when the state of the unmanned boat system is unmeasurable, an interval type-II Lumberg observer is established.

[0175] S32: Based on the interval type-2 fuzzy model and the interval type-2 Lumberg observer, an error system is constructed.

[0176] S33: According to the error evolution trajectory corresponding to the error system, the system state is expanded to obtain an interval type-II observer based on the observation error type-II augmented system.

[0177] In steps S31 to S33, first, when the state of the unmanned boat system is unmeasurable, the present invention designs an interval type II Lumberg observer based on the system output:

[0178]

[0179] Where L is the observer gain and the estimation error is defined as The gain of the observer needs to be designed offline, and then a specific error system can be constructed, expressed as:

[0180]

[0181] Based on the constructed error evolution trajectory and the original system state, the system state is expanded.

[0182] S4: Through the online state feedback predictive control algorithm and the dynamic output feedback predictive control algorithm, a dynamic robust model prediction is performed on the unmanned boat to achieve stabilization control of the unmanned boat.

[0183] In step S4, the theorem is first given: given a scalar β>0, the augmented interval type-2 fuzzy system is convergent under the action of the dynamic output feedback control rate and satisfies the control input constraint, state constraint and H ∞ Performance Index The premise is that there exists an observation matrix L and a positive definite matrix N x , N c The following optimization problem is solvable:

[0184]

[0185] At this time, the control gain

[0186] like Figure 5 As shown, step S4 further includes:

[0187] S41: Establish a type-II feedback control rate according to the estimated state of the system, substitute the type-II feedback control rate into the interval type-II observer, and obtain a basic dynamic output feedback system.

[0188] In steps S41 to S43, firstly, based on the estimated state of the system, a type II feedback control rate is designed:

[0189]

[0190] Substituting the type-II feedback control rate into the observation-error augmentation system, a unified dynamic output feedback dynamic system is formed:

[0191]

[0192] S42: Introducing the robust positive invariant set and the membership function information into the basic dynamic output feedback system to obtain an updated dynamic output feedback system.

[0193] In step S42, similar to steps S231 to S233, firstly, the robust positive invariant set is introduced, and the min-max problem of robust predictive control optimization in the infinite time domain with dynamic output feedback is converted into the min problem of minimizing the upper bound of the performance index, and the Lyapunov attenuation of the observation-error augmented system and the robust performance constraints are converted into matrix inequalities. Finally, the membership function vertex envelope method is used to introduce the membership function information into the augmented system inequality.

[0194] S43: Solving the updated dynamic output feedback system through an online state feedback predictive control algorithm and a corresponding dynamic output feedback predictive control algorithm to perform robust dynamic output feedback predictive control on the unmanned boat.

[0195] Finally, in step S43, by solving the optimization problem given in step S4, there exists an observation matrix L and a positive definite matrix N x , N c The optimization problem is solved, and the unmanned boat control can be performed by combining the unmanned boat dynamic output feedback robust model predictive control algorithm obtained in the aforementioned step S2.

[0196] Specifically, the system observation state and error state are initialized online, and then the optimization problem is solved to obtain the state feedback sub-gain K i =Y i (N x ) -1 And the observation matrix L, then combine these sub-gains with the membership functions (MFs) to obtain the local control law u(k). If it is not satisfied, u(k) = 0, and the control signal u(k) is input into the system, then k is updated to k+1 and the control signal u(k+1) is returned and input into the system, and it is continuously updated.

[0197] like Figures 6 to 9 FIG. 1 is a schematic diagram of the state of an unmanned boat in a specific application example of the present invention. Figure 6 To predict the actual state of the unmanned boat for state feedback, multiple curves x mn (m=1,2;n=1,2,…,6) represents the actual state of the system under different control rates, Figure 7 To output feedback to predict the actual state of the unmanned boat, it shows the robustness of the system when the state is unpredictable. n (n=1,2,…,6) represents the actual state of the system under different control rates; Figure 8 The output feedback predictive control of the unmanned boat observation state shows that when there is a disturbance, the unmanned boat observation system state can converge to the original system state. The multiple curves represent each The estimated status of Fig. 9The output feedback predictive control of the unmanned boat control input shows the control input convergence of dynamic output feedback, where multiple curves represent each u n (n=1,2,3) control input.

[0198] Figures 6 to 9 It can be seen that both the system state and the control input converge quickly. Although the disturbance input is large relative to the system state, both the state feedback robust control method algorithm corresponding to step S2 of the present invention and the output feedback robust control method algorithm corresponding to step S4 can effectively stabilize the system, and the output feedback can stabilize the unmanned boat faster due to its dynamic characteristics.

[0199] The present invention comprehensively considers the external disturbance, constraint and the unmeasurable state of the system, and proposes an optimization algorithm based on online dynamic output feedback robust predictive control, so that the unmanned boat can complete stabilization control under complex sea conditions. Compared with the traditional state-based control method without considering constraints, the present invention has the following beneficial effects: robust adaptation to complex environments: the designed online predictive control algorithm considers external disturbances such as ocean currents and waves, control input and state constraints, and the unmanned boat can still safely complete stabilization control even if there is disturbance; low conservatism: in the optimization problem, the membership function is used to unify the metric space, and the information of the interval type II membership function is introduced through the vertex envelope. The information reflects the nonlinear dynamic behavior of the system to a certain extent, which not only expands the range of the stabilization state of the unmanned boat, but also improves the dynamic performance of the system; real-time calculation: the original interval accumulation infinite time domain optimization solution is converted into a convex optimization problem constrained by linear matrix inequality, which can be solved in real time; economic efficiency: the designed dynamic output feedback online predictive control algorithm does not need to use speed sensors and position sensors to obtain the system state, and has high economic benefits.

[0200] 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 make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robust dynamic output feedback predictive control method for an unmanned boat, characterized in that: include: S1: Establish an interval type-2 fuzzy model of the unmanned boat; S2: Establish an optimization problem of online state feedback robust predictive control, and obtain an online state feedback predictive control algorithm by solving the optimization problem; S3: Based on the interval type-2 fuzzy model, an interval type-2 observer of the unmanned boat is established; S4: Based on the interval type-II observer, the online state feedback predictive control algorithm and the dynamic output feedback predictive control algorithm are used to perform dynamic robust model prediction on the unmanned boat to achieve stabilization control of the unmanned boat.

2. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 1, characterized in that: Step S1 further comprises: S11: Based on the surface motion characteristics of the unmanned boat, the state space model of the unmanned boat is established; S12: Based on the state space model, a basic type-2 fuzzy model is established; S13: Obtain an interval type-2 fuzzy model by discretizing the basic type-2 fuzzy model.

3. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 2, characterized in that: Step S11 further comprises: S111: Select the three-degree-of-freedom unmanned boat dynamics model as the basic model; S112: Based on the stabilization control problem of the unmanned boat, the centripetal force and the Coriolis force matrix of the basic model are ignored to obtain a simplified model; S113: Based on the correspondence between the state and position of the unmanned boat and the correspondence between the state and the navigation speed, the simplified model is converted into a state space model in the form of a state equation.

4. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 2, characterized in that: Step S12 further comprises: S121: Based on the range of the heading angle of the unmanned boat, define the second type of fuzzy rules of the unmanned boat; S122: define upper and lower bounds of membership weight functions to obtain a type II membership weight function; S123: Based on the type-2 fuzzy rules and the type-2 membership weight function, a basic type-2 fuzzy model is obtained.

5. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 1, characterized in that: Step S2 further comprises: S21: Establish the optimization problem of online state feedback robust predictive control; S22: Introducing a robust positive invariant set into the optimization problem to obtain an update problem; S23: A solution method for the update problem based on the robust positive invariant set and the membership function information is used to obtain an online state feedback predictive control algorithm.

6. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 5, characterized in that: Step S21 further includes: S211: Characterize the input control torque of the unmanned boat and the state constraint of the unmanned boat to obtain the input control torque and the state constraint; S212: Based on the non-parallel allocation compensation method, an interval type-II observer for the unmanned boat is established; S213: According to the input control torque, the state constraint, and the interval type-II observer, an infinite time domain unmanned boat stabilization control optimization problem is established to obtain an optimization problem.

7. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 5, characterized in that: Step S23 further comprises: S231: Convert Lyapunov attenuation and robust performance constraints into first matrix inequality and second matrix inequality respectively; S232: based on the second matrix inequality combined with the robust positive invariant set, according to the input control torque and the state constraint, the control input constraint and the state constraint are respectively converted into a third matrix inequality and a fourth matrix inequality; S233: Based on the membership function vertex envelope method, membership function information is introduced into the first matrix inequality, the third matrix inequality, and the fourth matrix inequality to obtain a membership matrix inequality; S234: Solving the update problem based on the membership matrix inequality to obtain an online state feedback predictive control algorithm.

8. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 7, characterized in that: The online state feedback predictive control algorithm in step S234 specifically includes: S2341: Initialize system status; S2342: Solve the update problem to obtain a state feedback sub-gain; S2343: combining the state feedback sub-gain with the membership function to obtain a local control law; S2344: Input the control signal corresponding to the local control law into the control system, and perform update control according to the time step.

9. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 1, characterized in that: Step S3 further comprises: S31: Based on the system output when the state of the unmanned boat system is unmeasurable, an interval type-II Lumberg observer is established; S32: constructing an error system based on the interval type-2 fuzzy model and the interval type-2 Lumberg observer; S33: According to the error evolution trajectory corresponding to the error system, the system state is expanded to obtain an interval type-II observer based on the observation error type-II augmented system.

10. The method for robust dynamic output feedback predictive control of an unmanned boat according to claim 1, characterized in that: Step S4 further comprises: S41: establishing a type-II feedback control rate according to the estimated state of the system, substituting the type-II feedback control rate into the interval type-II observer, and obtaining a basic dynamic output feedback system; S42: Introducing the robust positive invariant set and the membership function information into the basic dynamic output feedback system to obtain an updated dynamic output feedback system; S43: Solving the updated dynamic output feedback system through an online state feedback predictive control algorithm and a corresponding dynamic output feedback predictive control algorithm to perform robust dynamic output feedback predictive control on the unmanned boat.

Images