Control method and device for water surface multi-agent system

Abstract

The invention relates to a control method and device for a water surface multi-agent system, and belongs to the technical field of automatic driving of ships, and the method comprises the steps: taking the position and speed information of agents as state variables, and constructing a mathematical model of the water surface multi-agent system; according to a universal approximation principle of a fuzzy logic system, carrying out approximation on a nonlinear function in the mathematical model, and constructing a fuzzy state observer used for estimating an unmeasurable state in the mathematical model; and determining a generalized state error in which a time-varying formation function is introduced, and calculating to obtain target control input according to a sliding mode control algorithm. The unmeasurable information in the water surface multi-agent system is accurately estimated through the fuzzy state observer, the dependence of the controller on the speed state information of the single ship is reduced, the control precision is improved, meanwhile, the time-varying formation function and the sliding mode control algorithm are combined, the formation pattern of the water surface multi-agent system changes along with time, and the control precision is improved. The flexibility of the time-varying formation is higher.

Description

Technical Field

[0001] The present application relates to the field of ship autonomous driving technology, and in particular to a control method and device for a surface multi-agent system. Background Art

[0002] Ship motion has the characteristics of large time lag, large inertia, and nonlinearity. Changes in speed and loading lead to parameter perturbations in the control model. Changes in navigation conditions, interference from environmental parameters, and inaccurate measurements all cause uncertainty in the ship's heading control system. In the face of these problems caused by nonlinear uncertainty, intelligent algorithms have emerged and are constantly being applied to the field of ship heading control, such as adaptive control, robust control, fuzzy adaptive control, iterative sliding mode control, and minimum parameter learning methods.

[0003] In the prior art, the design of surface multi-agent systems often uses a state feedback control method, which is based on the assumption that all state information of the surface multi-agent system is known.

[0004] However, due to the physical limitations of sensors and premature measurement, the speed status information of a single ship in the multi-agent system cannot be measured in real time during actual navigation, resulting in the inability to transmit accurate speed status information to the controller for updating the control signal, resulting in poor control accuracy of the surface multi-agent system. Summary of the invention

[0005] In view of this, it is necessary to provide a control method and device for a surface multi-agent system to solve the problem of poor control accuracy in the prior art.

[0006] In order to solve the above problems, the present application provides a control method for a surface multi-agent system, comprising: The position and velocity information of the agent are used as state variables to construct the initial mathematical model of the surface multi-agent system; According to the universal approximation principle of fuzzy logic system, the nonlinear function in the initial mathematical model is approximated to obtain the target mathematical model; According to the input and output of the target mathematical model, a fuzzy state observer is constructed for estimating the unmeasurable state in the target mathematical model; Based on the target mathematical model and the fuzzy state observer, determining a generalized state error that introduces a time-varying formation function; According to the sliding mode control algorithm, the target control input of the surface multi-agent system is obtained based on the generalized state error calculation.

[0007] In some possible implementations, the initial mathematical model is:

[0008]

[0009] In the formula, Represents the first The position status information of each agent, Represents the first The speed state information of each agent, Represents the first The external disturbance to which an agent is subjected, Represents the first The target control input of an agent is represents a nonlinear function; The target mathematical model is:

[0010]

[0011] In the formula, Represents the first The optimal weight matrix for each agent is Represents the first The fuzzy basis function of each agent is Represents the estimated The speed state information of each agent, Represents the first The fuzzy minimum approximation error of the agents, Represents the first The approximation error of an agent.

[0012] In some possible implementations, the fuzzy state observer is:

[0013]

[0014] In the formula, Represents the estimated The position status information of each agent, Represents the estimated The speed state information of each agent, Represents the estimated The weight matrix of each agent, and Represents a constant.

[0015] In some possible implementations, the generalized state error is:

[0016]

[0017]

[0018]

[0019] In the formula, Represents the first The generalized state error of an agent, represents the time-varying formation function used to adjust the formation of the surface multi-agent system, Represents the first The agent and The quality of communication between agents, Represents the first The weighted value of the communication quality between an agent and all other agents, is an adjacency matrix, when When it represents a leader, Change to .

[0020] In some possible implementations, according to a sliding mode control algorithm, a target control input of the surface multi-agent system is obtained based on the generalized state error calculation, including: Based on the generalized state error, determining a time-varying sliding mode variable; Determining a time-varying sliding mode surface based on the time-varying sliding mode variable; Based on the time-varying sliding surface, a target control input is calculated.

[0021] In some possible implementations, the time-varying sliding mode variables include time-varying terminal sliding mode variables and time-varying linear auxiliary sliding mode variables, the time-varying sliding mode surface includes a time-varying terminal sliding mode surface and a time-varying linear auxiliary sliding mode surface, and the time-varying linear auxiliary sliding mode surface is used to switch from the time-varying terminal sliding mode surface to the time-varying linear auxiliary sliding mode surface when encountering a singularity problem.

[0022] In some possible implementations, the time-varying terminal sliding mode variable and the time-varying linear auxiliary sliding mode variable are respectively:

[0023]

[0024]

[0025]

[0026]

[0027] In the formula, Represents the first The time-varying terminal sliding mode variables of the agents, Represents the first The time-varying linear auxiliary sliding mode variables of the agents, Represents the first The generalized state error of an agent, represents a normal number, represents a constant, The diagonal elements are , and The matrix of The diagonal elements are , and The matrix of represents the symbolic function, , and Represents the first Agent generalized state error The 1st, 2nd and 3rd components of .

[0028] In some possible implementations, calculating a target control input based on the time-varying sliding surface includes: Based on the time-varying sliding surface, an initial control input is calculated; Based on the initial control input, a target control input after considering actuator saturation is calculated.

[0029] In some possible implementations, the method further includes: determining an actual control input based on the target control input, where the actual control input is:

[0030]

[0031]

[0032] In the formula, express The actual control input at the moment, express The target control input at the moment, represents the measurement error of the control input, express Target control input at time and Represents a positive integer.

[0033] The present application also provides an output feedback control device for a surface multi-agent system, comprising: A model building unit, used to build an initial mathematical model of the surface multi-agent system by taking the position and velocity information of the agent as state variables; A fuzzy observation unit, used to approximate the nonlinear function in the initial mathematical model according to the universal approximation principle of the fuzzy logic system to obtain a target mathematical model, and to construct a fuzzy state observer for estimating the unmeasurable state in the target mathematical model according to the input and output of the target mathematical model; A sliding mode control unit is used to determine the generalized state error introduced with the time-varying formation function based on the target mathematical model and the fuzzy state observer, and to obtain the target control input of the surface multi-agent system based on the generalized state error calculation according to the sliding mode control algorithm.

[0034] The beneficial effects of the present application are as follows: the control method for the surface multi-agent system provided by the present application uses a fuzzy state observer to accurately estimate the unmeasurable information in the surface multi-agent system, solves the output feedback problem of the multi-agent system, and effectively reduces the controller's dependence on the speed state information of a single ship. While solving problems such as uncertain multi-agent system models, unknown speed state information, input saturation, and uncertain external interference, the control accuracy is effectively improved.

[0035] Furthermore, the present application combines the time-varying formation function with the sliding mode control method and proposes a novel time-varying sliding mode control method, so that the formation of the surface multi-agent system can change over time. Compared with the traditional time-invariant formation, the flexibility of this time-varying formation is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic flow chart of an embodiment of a control method for a surface multi-agent system provided in the present application; Figure 2 For this application Figure 1 A schematic flow chart of an embodiment of step S105; Figure 3 For this application Figure 2 A schematic flow chart of an embodiment of step S203; Figure 4 A schematic diagram of the structure of an embodiment of an output feedback control device for a surface multi-agent system provided in the present application. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0038] It should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowcharts may be implemented out of order, and steps without logical contextual relationships may be reversed in order or implemented simultaneously. In addition, those skilled in the art may add one or more other operations to the flowcharts, or may remove one or more operations from the flowcharts, under the guidance of the content of this application. Some of the block diagrams shown in the accompanying drawings are functional entities, and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor systems and / or microcontroller systems.

[0039] The descriptions of "first", "second", etc. involved in the embodiments of the present application are only used to describe the implicit purpose, and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the technical features defined as "first" and "second" may explicitly or implicitly include at least one of the features. "And / or" describes the association relationship of the associated objects, indicating that three relationships may exist, for example: A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone.

[0040] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0041] The present application provides a control method for a surface multi-agent system, which is described below.

[0042] Figure 1 A flow chart of an embodiment of a control method for a surface multi-agent system provided in this application is as follows: Figure 1 As shown, the control method for the surface multi-agent system includes: S101, using the position and speed information of the agent as state variables to construct an initial mathematical model of the surface multi-agent system; In a specific embodiment, the collected single ship information is used to consider the nonlinear characteristics of the surface multi-agent system, and the initial mathematical model of the surface multi-agent system is constructed as follows:

[0043]

[0044] In the formula, Represents the position status information of the surface agent, represents the transformation matrix that transforms the hull coordinate system into the earth coordinate system, Represents the speed state information of the surface intelligent body in the hull coordinate system, represents the inertia matrix of the surface agent, represents the Coriolis centripetal force matrix of the surface agent, represents the damping matrix of the water surface agent, represents the external disturbance to the surface agent, Represents the control input of the surface agent.

[0045] make , the above initial mathematical model becomes:

[0046]

[0047] In the formula, Represents the velocity state information of the surface agent in the geodetic coordinate system, represents the Coriolis centripetal force matrix of the surface agent, represents the damping matrix of the water surface agent, represents the external disturbance to the surface agent, Represents the control input of the surface agent.

[0048] Re-order , the final initial mathematical model is:

[0049]

[0050] In the formula, Represents the first The position status information of each agent, Represents the first The speed state information of each agent, Represents the first The external disturbance to which an agent is subjected, Represents the first The target control input of an agent, Represents a nonlinear function, where the speed state information is unmeasurable.

[0051] S102, according to the universal approximation principle of the fuzzy logic system, approximate the nonlinear function in the initial mathematical model to obtain the target mathematical model; In a specific embodiment, it includes: First, define the ideal parameter vector of the surface multi-agent system as:

[0052] In the formula, express The tight set, express Tight set.

[0053] Then, according to the above ideal parameter vector, the fuzzy minimum approximation error can be obtained: for:

[0054] Then, according to the universal approximation principle of fuzzy logic system, nonlinear function can be approximated by a fuzzy logic system as:

[0055] In the formula, Represents the first The optimal weight matrix for each agent is Represents the first The fuzzy basis function of each agent is Represents the first The approximation error of an agent.

[0056] Substituting the above formula into the initial mathematical model, the target mathematical model is obtained as follows:

[0057]

[0058] In the formula, Represents the first The position status information of each agent, Represents the first The speed state information of each agent, Represents the first The external disturbance to which an agent is subjected, Represents the first The target control input of an agent is Represents the first The optimal weight matrix for each agent is Represents the first The fuzzy basis function of each agent is Represents the estimated The speed state information of each agent, Represents the first The fuzzy minimum approximation error of the agents, Represents the first The approximation error of an agent.

[0059] S103, constructing a fuzzy state observer for estimating an unmeasurable state in the target mathematical model according to the input and output of the target mathematical model; In a specific embodiment, the designed fuzzy state observer is:

[0060]

[0061] In the formula, Represents the estimated The position status information of each agent, Represents the estimated The speed state information of each agent, Represents the estimated The weight matrix of each agent, and Represents a constant.

[0062] Defining Observation Error for:

[0063] The observation error dynamics is:

[0064]

[0065] Calculate the adaptive fuzzy update rate of the surface multi-agent system as:

[0066] In the formula, and Represents the first The design parameters of the agents are positive constants.

[0067] S104, based on the target mathematical model and the fuzzy state observer, determining the generalized state error introduced into the time-varying formation function; It should be noted that in order to cope with more complex task requirements, a time-varying formation function for adjusting the formation of the surface multi-agent system is introduced into the generalized state error of the surface multi-agent system. Specifically, the first The agent and The state error of the agent about the position for:

[0068]

[0069] In the formula, when When it means leader, Change to , there is a leader in the surface multi-agent system, the position status information of the leader is known, and the mathematical model of the leader is:

[0070]

[0071] In the formula, Indicates the position status of the leader, Indicates the speed state of the leader, Represents an unknown bounded function.

[0072] In determining the surface multi-agent system The generalized state error of an agent for:

[0073]

[0074]

[0075]

[0076] In the formula, represents the time-varying formation function used to adjust the formation of the surface multi-agent system, Represents the first The agent and The quality of communication between agents, Represents the first The weighted value of the communication quality between an agent and all other agents, is an adjacency matrix, when When it represents a leader, Change to .

[0077] S105. According to the sliding mode control algorithm, the target control input of the surface multi-agent system is obtained based on the generalized state error calculation.

[0078] It should be noted that, considering the convergence problem of generalized state error, a time-varying sliding mode control method is adopted. By adjusting the time-varying function to change the formation, the finite time theory is applied to enable the surface multi-agent system to realize formation within a finite time.

[0079] Compared with the prior art, the present application uses a fuzzy state observer to accurately estimate the unmeasurable information in the surface multi-agent system, solves the output feedback problem of the multi-agent system, effectively reduces the controller's dependence on the speed state information of a single ship, and effectively improves the control accuracy while solving problems such as uncertain multi-agent system models, unknown speed state information, input saturation, and uncertain external interference.

[0080] Furthermore, the present application combines the time-varying formation function with the sliding mode control method and proposes a novel time-varying sliding mode control method, so that the formation of the surface multi-agent system can change over time. Compared with the traditional time-invariant formation, the flexibility of this time-varying formation is improved.

[0081] In order to better calculate the target control input, in some embodiments, such as Figure 2 As shown, step S105 specifically includes: S201, determining a time-varying sliding mode variable based on a generalized state error; S202, determining a time-varying sliding mode surface based on the time-varying sliding mode variable; S203: Calculate and obtain target control input based on the time-varying sliding surface.

[0082] Furthermore, considering the singularity problem of sliding mode control, in some embodiments, the system state space is divided into two regions, and a time-varying terminal sliding mode surface and a time-varying linear auxiliary sliding mode surface are designed respectively, as well as a corresponding time-varying terminal sliding mode controller and a time-varying linear auxiliary sliding mode controller. When encountering a singularity problem, the time-varying terminal sliding mode is switched to the time-varying linear auxiliary sliding mode to improve the control accuracy of the surface multi-agent system. Specifically, the time-varying terminal sliding mode variables and the time-varying linear auxiliary sliding mode variables are:

[0083]

[0084]

[0085]

[0086]

[0087] In the formula, Represents the first The time-varying terminal sliding mode variables of the agents, Represents the first The generalized state error of an agent, represents a normal number, represents a constant, The diagonal elements are , and The matrix of The diagonal elements are , and The matrix of represents the symbolic function, , and Represents the first Agent generalized state error The first, second and third components of Represents the first The time-varying linear auxiliary sliding mode variables of each agent.

[0088] The corresponding time-varying terminal sliding surface and time-varying linear auxiliary sliding surface are:

[0089]

[0090] In the formula, Represents the first The time-varying terminal sliding surface of the agent, Represents the first Time-varying linear auxiliary sliding surface for three intelligent agents.

[0091] Considering the actuator saturation problem, in some embodiments, such as Figure 3 As shown, step S203 specifically includes: S301, calculating and obtaining an initial control input based on a time-varying sliding surface; In a specific embodiment, the initial control input is:

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099] In the formula, represents the initial control input, represents the initial control input corresponding to the time-varying terminal sliding surface, represents the initial control input corresponding to the time-varying linear auxiliary sliding surface, Represents the first The observation error of the position state information of each agent, Represents the first The observation error of the velocity state information of each agent.

[0100] S302 : Based on the initial control input, a target control input after considering actuator saturation is calculated.

[0101] In a specific embodiment, the target control input is:

[0102] In the formula, represents the target control input after considering actuator saturation, represents the initial control input, represents the symbolic function, represents the saturation boundary.

[0103] Considering that the traditional fixed time sampling mechanism will waste a lot of network resources and communication bandwidth, in some embodiments, a dynamic event triggering mechanism is designed at the controller end of the surface intelligent body to reduce the update frequency of the control input without affecting the formation behavior and control target, so as to save the communication resources required for the control signal transmission. Specifically, the method also includes: determining the actual control input based on the target control input, and the actual control input is:

[0104]

[0105]

[0106] In the formula, express The actual control input at the moment, express The target control input at the moment, Indicates the triggering time, represents the measurement error of the control input, express Target control input at time and Represents a positive integer.

[0107] Therefore, the controller is always calculating the target control input at each moment. Only at the trigger moment does the controller pass the target control input to the actuator, and the actuator updates the actual control input. Whether the actual control input is updated depends on the measurement error of the control input. Before the next event triggers, the actual control input remains the target control input at the last trigger moment. .

[0108] In order to better implement a control method for a surface multi-agent system in an embodiment of the present application, based on a control method for a surface multi-agent system, correspondingly, Figure 4 As shown, the embodiment of the present application also provides an output feedback control device 400 for a surface multi-agent system, comprising: A model building unit 401 is used to build an initial mathematical model of the surface multi-agent system by taking the position and speed information of the agent as state variables; A fuzzy observation unit 402 is used to approximate the nonlinear function in the initial mathematical model according to the universal approximation principle of the fuzzy logic system to obtain a target mathematical model, and to construct a fuzzy state observer for estimating the unmeasurable state in the target mathematical model according to the input and output of the target mathematical model; The sliding mode control unit 403 is used to determine the generalized state error introduced with the time-varying formation function based on the target mathematical model and the fuzzy state observer, and obtain the target control input of the surface multi-agent system based on the generalized state error calculation according to the sliding mode control algorithm.

[0109] The output feedback control device 400 for the surface multi-agent system provided in the above embodiment can implement the technical solution described in the above embodiment of the control method for the surface multi-agent system. The specific implementation principles of the above units can refer to the corresponding contents in the above embodiment of the control method for the surface multi-agent system, which will not be repeated here.

[0110] The above is a detailed introduction to a control method for a surface multi-agent system provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, based on the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

[0111] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed in the present application should be covered within the protection scope of the present application.

Claims

1. A control method for a surface multi-agent system, characterized in that: include: The position and velocity information of the agent are used as state variables to construct the initial mathematical model of the surface multi-agent system; According to the universal approximation principle of the fuzzy logic system, the nonlinear function in the initial mathematical model is approximated to obtain the target mathematical model; According to the input and output of the target mathematical model, a fuzzy state observer is constructed for estimating the unmeasurable state in the target mathematical model; Based on the target mathematical model and the fuzzy state observer, determining a generalized state error that introduces a time-varying formation function; According to the sliding mode control algorithm, the target control input of the surface multi-agent system is calculated based on the generalized state error.

2. The control method for a surface multi-agent system according to claim 1, characterized in that: The initial mathematical model is: In the formula, Represents the first The position status information of each agent, Represents the first The speed state information of each agent, Represents the first The external disturbance to which an agent is subjected, Represents the first The target control input of an agent is represents a nonlinear function; The target mathematical model is: In the formula, Represents the first The optimal weight matrix for each agent is Represents the first The fuzzy basis function of each agent is Represents the estimated The speed state information of each agent, Represents the first The fuzzy minimum approximation error of the agents, Represents the first The approximation error of an agent.

3. The control method for a surface multi-agent system according to claim 2, characterized in that: The fuzzy state observer is: In the formula, Represents the estimated The position status information of each agent, Represents the estimated The speed state information of each agent, Represents the estimated The weight matrix of each agent, and Represents a constant.

4. The control method for a surface multi-agent system according to claim 3, characterized in that: The generalized state error is: In the formula, Represents the first The generalized state error of an agent, represents the time-varying formation function used to adjust the formation of the surface multi-agent system, Represents the first The agent and The quality of communication between agents, Represents the first The weighted value of the communication quality between an agent and all other agents, is an adjacency matrix, when When it represents a leader, Change to .

5. The control method for a surface multi-agent system according to claim 4, characterized in that: According to the sliding mode control algorithm, the target control input of the surface multi-agent system is obtained based on the generalized state error calculation, including: Based on the generalized state error, determining a time-varying sliding mode variable; Determining a time-varying sliding mode surface based on the time-varying sliding mode variable; Based on the time-varying sliding surface, a target control input is calculated.

6. The control method for a surface multi-agent system according to claim 5, characterized in that: The time-varying sliding mode variables include time-varying terminal sliding mode variables and time-varying linear auxiliary sliding mode variables, the time-varying sliding mode surface includes a time-varying terminal sliding mode surface and a time-varying linear auxiliary sliding mode surface, and the time-varying linear auxiliary sliding mode surface is used to switch from the time-varying terminal sliding mode surface to the time-varying linear auxiliary sliding mode surface when encountering a singularity problem.

7. The control method for a surface multi-agent system according to claim 6, characterized in that: The time-varying terminal sliding mode variable and the time-varying linear auxiliary sliding mode variable are respectively: In the formula, Represents the first The time-varying terminal sliding mode variables of the agents, Represents the first The time-varying linear auxiliary sliding mode variables of the agents, Represents the first The generalized state error of an agent, represents a normal number, represents a constant, The diagonal elements are , and The matrix of The diagonal elements are , and The matrix of represents the symbolic function, , and Represents the first Agent generalized state error The 1st, 2nd and 3rd components of .

8. The control method for a surface multi-agent system according to claim 7, characterized in that: Based on the time-varying sliding surface, the target control input is calculated, including: Based on the time-varying sliding surface, an initial control input is calculated; Based on the initial control input, a target control input after considering actuator saturation is calculated.

9. The control method for a surface multi-agent system according to claim 8, characterized in that: The method further includes: determining an actual control input based on the target control input, wherein the actual control input is: In the formula, express The actual control input at the moment, express The target control input at the moment, represents the measurement error of the control input, express The target control input at the moment, and Represents a positive integer.

10. An output feedback control device for a surface multi-agent system, characterized in that: include: A model building unit, used to build an initial mathematical model of the surface multi-agent system by taking the position and velocity information of the agent as state variables; A fuzzy observation unit, used to approximate the nonlinear function in the initial mathematical model according to the universal approximation principle of the fuzzy logic system to obtain a target mathematical model, and to construct a fuzzy state observer for estimating the unmeasurable state in the target mathematical model according to the input and output of the target mathematical model; A sliding mode control unit is used to determine the generalized state error introduced with the time-varying formation function based on the target mathematical model and the fuzzy state observer, and to obtain the target control input of the surface multi-agent system based on the generalized state error calculation according to the sliding mode control algorithm.

Images