A multi-agent cooperative control method and medium based on hybrid trigger mechanism

Abstract

The present invention discloses a multi-agent cooperative control method based on a hybrid trigger mechanism, including: determining the agent dynamics model of the multi-agent system; determining the network topology model of the multi-agent system; determining a distributed cooperative control protocol according to the agent dynamics model and the network topology model; determining an asynchronous time trigger mechanism according to the network topology model and the distributed cooperative control protocol; determining a distributed event trigger mechanism according to the network topology model and the distributed cooperative control protocol; according to the above-mentioned protocol, the self-controller updates the sampled agent relative output and the sampled neighbor controller state, and calculates the agent's own input in real time, so that the multi-agent system can achieve cooperation. It constructs a distributed cooperative control protocol based on the relative output information between agents, without measuring the agent's own absolute output information, and the agents only need to interact with each other intermittently, asynchronously, and distributedly, which has good practical value.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-agent collaborative control, and relates to a multi-agent collaborative control method and medium based on a hybrid trigger mechanism, specifically a multi-agent collaborative control method and medium based on a hybrid time / event trigger mechanism. Background Art

[0002] Multi-agent cooperative control refers to the technology of driving all agents to cooperate and achieve the predetermined goal by designing appropriate coordinated control strategies for a dynamic network system composed of multiple agents. Due to its extremely important application value in the fields of industry, energy, military, aerospace, etc., multi-agent cooperative control technology has become one of the research hotspots in related fields in the past two decades, and with the development of information technology, it has played an increasingly important role. In industry, by modeling production machines as intelligent entities and using information technology to connect them into an organic whole, multi-agent collaborative control technology can be used to improve the intelligence level of factories; in logistics, through the collaboration between logistics unmanned vehicles and / or drones, multi-agent collaborative control technology can be used to improve the efficiency of logistics path planning and reduce logistics costs; in energy, multi-agent collaborative control technology can better coordinate energy networks to achieve power generation / consumption balance and improve energy utilization efficiency through information sharing between various power generation units and power consumption units; in the military, aerospace, aviation, navigation and other fields, multi-agent collaborative control technology has become the basic means for various carriers to achieve unmanned, clustered and intelligent operations, and can give or enhance the capabilities of multi-carrier systems in perception, navigation, execution, survival, etc., and expand functions that single carriers do not have to complete more complex tasks.

[0003] There are relatively systematic and mature theories and methods for multi-agent cooperative control. Taking the most representative consistency goal as an example, the existing cooperative control design methods have effectively solved the consistency control problems under various intelligent agent dynamic models such as integral multi-agent systems, linear multi-agent systems, and nonlinear multi-agent systems.

[0004] Regardless of the collaborative control method adopted, the configuration of the multi-agent system requires the controller to first be able to obtain the state or output information between adjacent agents, and then design a collaborative control strategy based on this information.

[0005] Since the classic collaborative control method requires the controller to continuously obtain information about neighboring agents, it has disadvantages such as heavy communication load and low communication efficiency in practical applications.

[0006] On the one hand, information exchange between agents requires the help of communication networks, but the distributed nature of multi-agent networks (such as mobile robots, underwater unmanned boats, etc.) makes it difficult to maintain continuous communication; on the other hand, the communication capabilities and resources of a single agent are often very limited, and continuous information exchange will consume too much network channel resources and reduce the survivability of the multi-agent system.

[0007] In recent years, with the help of the idea of ​​event triggering, the multi-agent cooperative control method based on event triggering has attracted widespread attention at home and abroad and has become a research hotspot in the field of multi-agent cooperative control. In multi-agent cooperative control, the role of the existing event triggering mechanism is usually to determine the sampling time of the agent's own state. When the sampling error exceeds a certain set threshold, the agent samples its own state and distributes the sampled state to the neighboring agent. Therefore, this type of event triggering mechanism requires the agent to be able to obtain its own absolute information. In order to meet the needs of cooperative control, this requires all agents to measure information based on a common global coordinate system, which puts extremely high demands on the perception ability of the agent. For example, in some typical application scenarios such as deep space cooperative exploration and indoor navigation without GPS signals, it is extremely difficult to establish an accurate common coordinate system for all agents. In addition, considering that the event triggering conditions need to be constantly checked, the event triggering method applied to the multi-agent system should also have an inherent working mechanism of spatial distribution and time asynchrony. In summary, the existing multi-agent cooperative control method based on event triggering still has great limitations, and it is very necessary to provide a more practical multi-agent cooperative control design method for multi-agent cooperative control. Summary of the invention

[0008] In view of the defects in the prior art, the purpose of the present invention is to provide a multi-agent collaborative control method and medium based on a hybrid trigger mechanism.

[0009] According to one aspect of the present invention, a multi-agent collaborative control method based on a hybrid time / event trigger mechanism is provided, comprising:

[0010] Determine the agent dynamics model of the multi-agent system;

[0011] Determine the network topology model of the multi-agent system;

[0012] Determining a distributed collaborative control protocol according to the intelligent agent dynamics model and the network topology model;

[0013] Determining an asynchronous time trigger mechanism according to the network topology model and the distributed collaborative control protocol;

[0014] Determining a distributed event triggering mechanism according to the network topology model and the distributed collaborative control protocol;

[0015] According to the distributed collaborative control protocol, the asynchronous time trigger mechanism and the distributed event trigger mechanism, the self-controller intermittently, asynchronously and distributedly updates the sampled agent relative output and the sampled neighbor controller status, and calculates the agent's own input in real time, so that the multi-agent system can achieve collaboration.

[0016] Preferably, the agent dynamics model of the multi-agent system includes:

[0017] The multi-agent system includes N agents, and the dynamic model of the i-th agent is:

[0018]

[0019]

[0020] The first equation is the state equation of the i-th agent, and the second equation is the output equation of the i-th agent; and are the state function and output function of the ith agent respectively; is the state of the ith agent, n x is the state dimension, For x i The time derivative of (t); is the self-input of the ith agent, n u is the input dimension; is the self-output of the ith agent, n y is the output dimension; t represents the time variable.

[0021] Preferably, the state function of the i-th agent is

[0022]

[0023] The output function of the i-th agent is

[0024]

[0025] in and is the parameter matrix of the ith agent, which is determined by the dynamic characteristics of the agent itself.

[0026] Preferably, the network topology model includes nodes, edges and edge weights, wherein:

[0027] The one node represents an agent;

[0028] The edges represent the connection relationship between the agents;

[0029] The edge weight represents the weight of information transmitted through the edge;

[0030] The neighboring agent refers to a node that has an edge connected to the own agent in the network topology model.

[0031] Preferably, use Represent the network topology model, using represents a node set, where N is the number of agents, represents the edge set, and the neighbor set of node i is expressed as

[0032] The figure The adjacency matrix of is defined as The Laplacian matrix is ​​defined as where a ij ≥0 indicates the weight of the edge (i, j), the matrix as well as

[0033] Preferably, the distributed collaborative control protocol is:

[0034]

[0035]

[0036] in is the state of the controller of the i-th agent, n x is the state dimension; is the self-input of the ith agent, n u is the input dimension; and are the state function and output function of the controller of the i-th agent respectively; represents the controller state of the first type of sampling, defined as Represents the controller state for the second type of sampling, defined as is the first sampling time sequence, is the second sampling time sequence; represents the relative output of the sampled agent associated with edge (j, i), y i is the self-output of the ith agent.

[0037] Preferably, the state function of the self-controller of the i-th agent is:

[0038]

[0039] The output function of the controller of the i-th agent is:

[0040]

[0041] Wherein, A, B and C are the parameter matrices of the agent dynamics model; is the first gain matrix of the control protocol, satisfying K = -B T P, where To satisfy the Riccati equation A T P+PA-PBB T Symmetric positive definite solution of P+Q=0, is any given symmetric positive definite matrix; is the second gain matrix of the control protocol, making the matrix A+FC Hurwitz stable; c is a positive constant that satisfies λ 2 For the figure The Laplace matrix of The smallest non-zero eigenvalue of the real part, Re(λ 2 ) is λ 2 The real part of .

[0042] Preferably, the asynchronous time trigger mechanism is:

[0043]

[0044] Where T i >0 is a constant;

[0045] The first sampling time sequence The selection needs to satisfy the asynchronous time trigger mechanism;

[0046] When the system time t reaches the sampling time The relative output of the agent sampled by the i-th agent update At the same time, the i-th agent updates its controller state of the first type of sampling

[0047] Preferably, the distributed event triggering mechanism is:

[0048]

[0049] in is the sampling error of its own controller state, Trigger function for the event;

[0050] The second sampling time sequence Determined by the distributed event triggering mechanism;

[0051] When the event trigger function satisfies f i When (·)≥0, the i-th agent updates its controller state of the second type of sampling At the same time, the updated controller state of the second type of sampling is broadcast to neighboring agents

[0052] When the event triggering function of the jth neighboring agent of the ith agent satisfies f j When (·)≥0, the jth neighbor agent of the ith agent updates its controller state of the second type of sampling At the same time, the i-th agent receives the updated controller state of the second type of sampling broadcasted by the j-th neighbor agent

[0053] According to a second aspect of the present invention, a computer-readable storage medium is provided, wherein the storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded by a processor and executes any one of the methods described.

[0054] Compared with the prior art, the technical solution disclosed in the present invention has the following beneficial effects:

[0055] An embodiment of the present invention provides a multi-agent collaborative control method based on a hybrid time / event trigger mechanism. The method constructs a distributed collaborative control protocol based on the relative output information between the agents. It does not need to measure the absolute output information of the agents themselves, avoids dependence on a common global coordinate system, and reduces the requirements of the collaborative control protocol on the perception capabilities of the agents. The method has great practical value.

[0056] Another embodiment of the present invention provides a multi-agent collaborative control method based on a hybrid time / event trigger mechanism. A time trigger mechanism and an event trigger mechanism are designed respectively for the relative output of the agents and the status of neighbor controllers, thereby realizing intermittent, distributed, and asynchronous information interaction between agents. There is no need to continuously monitor neighbor agents, which saves the perception and communication resources of the multi-agent system, reduces the difficulty of implementing the control protocol, reduces the risk of communication congestion and network attacks, and extends the life cycle of the actual multi-agent system. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0058] Figure 1 A flowchart of a multi-agent collaborative control method based on a hybrid time / event trigger mechanism provided in one embodiment of the present invention;

[0059] Figure 2 A system framework diagram of a multi-agent collaborative control method based on a hybrid time / event trigger mechanism provided by an embodiment of the present invention;

[0060] Figure 3 A network topology diagram of 10 intelligent agents provided in a preferred embodiment of the present invention;

[0061] Figure 4 A 0-50s agent state trajectory diagram provided by a preferred embodiment of the present invention;

[0062] Figure 5 A sampling signal trajectory diagram of a 0-50s distributed event trigger mechanism provided by a preferred embodiment of the present invention;

[0063] Figure 6 A sampling signal trajectory diagram of a 0-50s asynchronous time trigger mechanism provided by a preferred embodiment of the present invention;

[0064] Figure 7 This is an event interval diagram of a 0-50s distributed event triggering mechanism provided by a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0065] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several variations and improvements may be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0066] In order to facilitate understanding of the embodiments of this part, a brief introduction to the basic framework of the present invention is now given. The present application is a system composed of multiple agents. Each agent includes a dynamic model of the agent and a self-controller. Among them, each self-controller contains two different sampled controller states, namely, the first type of sampled controller state and the second type of sampled controller state.

[0067] Based on the above framework structure, the present invention provides an embodiment. Figure 1 is a flow chart of the multi-agent collaborative control method in this embodiment, Figure 2 For Figure 1 The corresponding system framework diagram of the multi-agent collaborative control method. Figure 1 and Figure 2 , the multi-agent collaborative control method based on the hybrid time / event trigger mechanism in this embodiment includes:

[0068] S1, determine the agent dynamics model of the multi-agent system;

[0069] S2, determine the network topology model of the multi-agent system;

[0070] S3, determining the distributed cooperative control protocol according to the agent dynamics model in S1 and the network topology model in S2;

[0071] S4, determining an asynchronous time triggering mechanism according to the network topology model in S2 and the distributed cooperative control protocol in S3;

[0072] S5, determining a distributed event triggering mechanism according to the network topology model in S2 and the distributed collaborative control protocol in S3;

[0073] S6, according to the distributed collaborative control protocol in S3, the asynchronous time trigger mechanism in S4 and the distributed event trigger mechanism in S5, the self-controller intermittently, distributedly and asynchronously updates the sampled agent relative output and the sampled neighbor controller status, and calculates the self-input of each agent in real time, so that the multi-agent system can finally achieve collaboration.

[0074] This embodiment avoids the controller's reliance on continuous time information between intelligent agents, improves the communication efficiency between intelligent agents, and overcomes the technical difficulty of low practicality caused by the high requirements of the existing methods on the perception ability of the intelligent agents.

[0075] In a preferred embodiment of the present invention, S1 is implemented: determining an agent dynamics model of a multi-agent system, which is used to characterize the dynamic properties satisfied by the agent. The agent dynamics model includes its own state, its own input, its own output and the mathematical relationship they satisfy.

[0076] Agent dynamics model of multi-agent system, including:

[0077] The multi-agent system consists of N agents, and the dynamic model of the i-th agent is

[0078]

[0079]

[0080] The first equation is the state equation of the i-th agent, and the second equation is the output equation of the i-th agent; and are the state function and output function of the ith agent respectively; is the state of the ith agent, n x is the state dimension, For x i The time derivative of (t); is the self-input of the ith agent, n uis the input dimension; is the self-output of the ith agent, n y is the output dimension; t represents the time variable. Specifically,

[0081] The state function of the i-th agent is:

[0082]

[0083] The output function of the i-th agent is:

[0084]

[0085] in and is the parameter matrix of the ith agent, which is determined by the dynamic characteristics of the agent itself.

[0086] In a preferred embodiment of the present invention, S2 is implemented to determine a network topology model of a multi-agent system, which is used to characterize the information interaction relationship between the own agent and the neighboring agents. The network topology model includes nodes, edges, and edge weights, where a node represents an agent, an edge represents the connection relationship between agents, and an edge weight represents the weight of information transmitted through the edge; a neighboring agent refers to a node in the network topology model that has an edge with the own agent.

[0087] Specifically, the network topology model is shown in Figure Indicates. represents a node set, where N is the number of agents, represents the edge set, and the neighbor set of node i is expressed as picture The adjacency matrix is ​​defined as The Laplacian matrix is ​​defined as where a ij ≥0 indicates the weight of the edge (i, j), the matrix as well as

[0088] In a preferred embodiment of the present invention, based on the agent dynamics model in the above embodiment S1 and the network topology model in the above embodiment S2, S3 is implemented to determine a distributed cooperative control protocol. The distributed cooperative control protocol is used to calculate the agent's own input. In the distributed cooperative control protocol, the information required by the self controller and related to the neighboring agents includes the sampled agent relative output and the controller state of the second type of samples of the neighbors.

[0089] Specifically, according to the agent dynamics model, the following Riccati equation is solved:

[0090] A TP+PA-PBB T P+Q=0,

[0091] in Any given symmetric positive definite matrix, A and B are the parameter matrices of the agent dynamics model, the matrix is a symmetric positive definite solution to the Riccati equation.

[0092] According to the network topology model, define λ i , For the picture The Laplace matrix of The eigenvalues ​​of are sorted to satisfy 0 = λ 1 <Re(λ 2 )≤…≤Re(λ N ), where λ 2 For the picture The Laplace matrix of The smallest non-zero eigenvalue of the real part, Re(λ i ) is λ i The real part of λ 2 The limitation of eigenvalues ​​makes it more universal.

[0093] definition and are two different sampling time sequences.

[0094] The distributed cooperative control protocol is designed as follows:

[0095]

[0096]

[0097] in is the state of the controller of the i-th agent, n x is the state dimension; is the self-input of the ith agent, n u is the input dimension; represents the controller state of the first type of sampling, defined as Represents the controller state for the second type of sampling, defined as is the first sampling time sequence, is the second sampling time sequence; represents the relative output of the sampled agent associated with edge (j, i), y i is the self-output of the ith agent; A, B and C are the parameter matrices of the agent dynamics model; is the first gain matrix of the control protocol, satisfying K = -B TP, where To satisfy the symmetric positive definite solution of the Riccati equation; is the second gain matrix of the control protocol, making the matrix A+FC Hurwitz stable; c is a positive constant that satisfies λ 2 For the picture The Laplace matrix of The smallest non-zero eigenvalue of the real part, Re(λ 2 ) is λ 2 The real part of .

[0098] The distributed cooperative control protocol in this embodiment can be used to enable the self-controller to intermittently update the sampled agent relative output and the sampled controller state.

[0099] In a preferred embodiment of the present invention, based on the network topology model in S2 and the distributed cooperative control protocol in S3 in the above embodiment, S4 is implemented to determine an asynchronous time trigger mechanism. The asynchronous time trigger mechanism is used to determine whether the agent resamples the relative output of the agent and whether to update the controller state of the first type of sampling of the agent. The asynchronous time trigger mechanism determines its next sampling time according to the last sampling time triggered by itself and the current system time.

[0100] Let the constant T i satisfy where ρ(·) represents the matrix spectral radius operator, and the constant T i The sampling period of the agent's relative output used to control sampling.

[0101] The asynchronous time trigger mechanism is designed as follows:

[0102]

[0103] The constant T i Indicates the sampling period.

[0104] When the system time t reaches the sampling time The relative output of the agent sampled by the i-th agent update At the same time, the i-th agent updates its controller state of the first type of sampling

[0105] The asynchronous time trigger mechanism in this embodiment can be used to enable the self-controller to asynchronously update the sampled agent relative output and the self-controller state of the first type of sampling.

[0106] In a preferred embodiment of the present invention, based on the network topology model in S2 and the distributed cooperative control protocol in S3 in the above embodiment, S5 is implemented to determine a distributed event triggering mechanism. The distributed event triggering mechanism is used to determine whether the own agent updates its own second-class sampled controller state and whether to broadcast the updated second-class sampled controller state to neighboring agents; the distributed event triggering mechanism determines its own next sampling time according to the sampling error of its own second-class sampled controller state, its own and neighboring second-class sampled controller states and the current system time.

[0107] The distributed event triggering mechanism is designed as follows:

[0108]

[0109] where f i (t) is the event trigger function, which is designed as:

[0110]

[0111] in is the sampling error of the controller state of the second type of sampling itself, l ii For the picture The Laplace matrix of The i-th diagonal element of ij is the weight of the edge (i, j), θ i (t) is a non-negative function. When the event trigger function satisfies f i When (·)≥0, the i-th agent updates its controller state of the second type of sampling At the same time, the updated controller state of the second type of sampling is broadcast to neighboring agents When the event triggering function of the jth neighboring agent of the ith agent satisfies f j When (·)≥0, the jth neighbor agent of the ith agent updates its controller state of the second type of sampling At the same time, the i-th agent receives the updated controller state of the second type of sampling broadcasted by the j-th neighbor agent

[0112] The distributed cooperative control protocol in this embodiment can be used to enable the self-controller to update the self-controller state of the second type of sampling and the neighbor controller state of the second type of sampling in a distributed manner.

[0113] A simulation experiment is given below to verify the feasibility of the multi-agent collaborative control method provided by the present invention.

[0114] Consider a multi-agent system consisting of 10 drones. The dynamic model of the agent is:

[0115]

[0116]

[0117] in C = [1 0], state x i The two components of represent the position and speed of the drone, and the output y i is the position of the drone. Assume that the cooperative control goal of the multi-agent system is to make each agent finally gather together and move at the same speed.

[0118] Consider Figure 3 The network topology model shown in the figure, where the weight of each edge is 1, then the minimum non-zero eigenvalue of the Laplace matrix is ​​λ 2 =0.3961.

[0119] Let Q = I, then determine the symmetric positive definite matrix Then determine the first gain matrix of the control protocol as K = -B T P = [-1 -1.7321]; the second gain matrix of the control protocol is selected as Make the matrix A+FC Hurwitz stable; choose the constant c as

[0120] Select the constant T in the asynchronous time trigger mechanism i for Satisfies the inequality:

[0121]

[0122] Select the non-negative function of the distributed event trigger mechanism as θ i (t) = exp(-0.4t),

[0123] Run the simulation experiment for 50 seconds. Figure 4 Give the agent state trajectory diagram from 0 to 50 seconds, Figure 5 The sampling signal trajectory diagram of the distributed event trigger mechanism from 0 to 50 seconds is given. Figure 6 The sampling signal trajectory diagram of the asynchronous time trigger mechanism of 0-50s is given. Figure 7 The event interval diagram of the distributed event triggering mechanism of 0-50s is given. At time t = 0s, each agent has a different initial position and initial movement speed, and when time t ≥ 10s, each agent gradually converges and continues to move at the same speed, verifying that the designed cooperative control protocol achieves the predetermined cooperative control objectives (such as Figure 4The traces of the sampled control signals show that the agents only need to exchange information intermittently and asynchronously, which reduces the communication load of the control protocol (e.g. Figure 5 and Figure 6 The event interval of the distributed event triggering mechanism is defined as The experimental results show that the event intervals of each agent In the range of 0-50s, it is strictly greater than zero, indicating that the designed distributed event trigger mechanism will not cause infinitely fast sampling action and has good practicality (such as Figure 7 shown).

[0124] In other embodiments of the present invention, there is also a computer-readable storage medium, in which at least one instruction, at least one program, code set or instruction set is stored, and the at least one instruction, the at least one program, the code set or instruction set is loaded by a processor and executes the method described in any one of the embodiments.

[0125] It should be noted that the steps in the method provided by the present invention can be implemented by using corresponding modules, devices, units, etc. in the system. Those skilled in the art can refer to the technical solution of the system to implement the step flow of the method, that is, the embodiments in the system can be understood as preferred examples for implementing the method, which will not be elaborated here.

[0126] Those skilled in the art know that, in addition to implementing the system and its various devices provided by the present invention in a purely computer-readable program code, the system and its various devices provided by the present invention can be made to implement the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices provided by the present invention can be considered as a hardware component, and the devices for implementing various functions included therein can also be considered as structures within the hardware component; the devices for implementing various functions can also be considered as both software modules for implementing the method and structures within the hardware component.

[0127] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various modifications or variations within the scope of the claims, which does not affect the essence of the present invention. The above preferred features can be used in any combination without conflicting with each other.

Claims

1. A multi-agent collaborative control method based on a hybrid time / event trigger mechanism, characterized in that: include: Determine the agent dynamics model of the multi-agent system; Determine the network topology model of the multi-agent system; Determining a distributed collaborative control protocol according to the intelligent agent dynamics model and the network topology model; Determining an asynchronous time trigger mechanism according to the network topology model and the distributed collaborative control protocol; Determining a distributed event triggering mechanism according to the network topology model and the distributed collaborative control protocol; According to the distributed collaborative control protocol, the asynchronous time trigger mechanism and the distributed event trigger mechanism, the self-controller intermittently, asynchronously and distributedly updates the sampled agent relative output and the sampled neighbor controller state, and calculates the agent's own input in real time, so that the multi-agent system can achieve collaboration; The distributed collaborative control protocol is: in is the state of the controller of the i-th agent, n x is the state dimension; is the self-input of the ith agent, n u is the input dimension; and are the state function and output function of the controller of the i-th agent respectively; represents the controller state of the first type of sampling, defined as Represents the controller state for the second type of sampling, defined as is the first sampling time sequence, is the second sampling time sequence; represents the relative output of the sampled agent associated with the edge (j, i) in the network topology model, y i is the self-output of the ith agent; The state function of the controller of the i-th agent is: The output function of the controller of the i-th agent is: Wherein, A, B and C are the parameter matrices of the agent dynamics model; is the first gain matrix of the control protocol, satisfying K = -B T P, where To satisfy the Riccati equation A T P+PA-PBB T Symmetric positive definite solution of P+Q=0, is any given symmetric positive definite matrix; is the second gain matrix of the control protocol, making the matrix A+FC Hurwitz stable; c is a positive constant that satisfies λ2 is the graph The Laplace matrix of The smallest non-zero eigenvalue of the real part, Re(λ2) is the real part of λ2; The network topology model includes nodes, edges and edge weights, wherein: One of the nodes represents an agent; The edges represent the connection relationship between the agents; The edge weight represents the weight of information transmitted through the edge; The neighboring agent refers to a node that has an edge with the own agent in the network topology model; Use pictures Represent the network topology model, using represents a node set, where N is the number of agents, represents the edge set, and the neighbor set of node i is expressed as The figure The adjacency matrix is ​​defined as The Laplacian matrix is ​​defined as where a ij ≥0 indicates the weight of the edge (i, j), the matrix as well as 2. According to claim 1, a multi-agent collaborative control method based on a hybrid time / event trigger mechanism is characterized in that: The agent dynamics model of the multi-agent system includes: The multi-agent system includes N agents, and the dynamic model of the i-th agent is: y i (t)=p(x i (t),u i (t),t), The first equation is the state equation of the i-th agent, and the second equation is the output equation of the i-th agent; and are the state function and output function of the ith agent respectively; is the state of the ith agent, n x is the state dimension, For x i The time derivative of (t); is the self-input of the ith agent, n u is the input dimension; is the self-output of the ith agent, n y is the output dimension; t represents the time variable.

3. The multi-agent collaborative control method based on a hybrid time / event trigger mechanism according to claim 2 is characterized in that: The state function of the ith agent is: The output function of the ith agent is: y i (t)=Cx i (t), in and is the parameter matrix of the ith agent, which is determined by the dynamic characteristics of the agent itself.

4. The multi-agent cooperative control method based on a hybrid time / event trigger mechanism according to claim 1 is characterized in that: The asynchronous time trigger mechanism is: Where T i >0 is a constant; The first sampling time sequence The selection needs to satisfy the asynchronous time trigger mechanism; When the system time t reaches the sampling time The relative output of the agent sampled by the i-th agent update At the same time, the i-th agent updates its controller state of the first type of sampling 5. The multi-agent cooperative control method based on a hybrid time / event trigger mechanism according to claim 1 is characterized in that: The distributed event triggering mechanism is: in is the sampling error of the controller state for the second type of sampling, Trigger function for the event; The second sampling time sequence Determined by the distributed event triggering mechanism; When the event trigger function satisfies f i When (·)≥0, the i-th agent updates its controller state of the second type of sampling At the same time, the updated controller state of the second type of sampling is broadcast to neighboring agents When the event triggering function of the jth neighboring agent of the ith agent satisfies f j When (·)≥0, the jth neighbor agent of the ith agent updates its controller state of the second type of sampling At the same time, the i-th agent receives the updated controller state of the second type of sampling broadcasted by the j-th neighbor agent 6. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded by the processor and executes the method according to any one of claims 1 to 5.

Images