A fully distributed optimization method and device for heterogeneous multi-intelligent systems under preset time

By acquiring agent information and determining directed communication topology information, using distributed optimization algorithms to calculate the minimized global cost function solution, the problem of low control efficiency in heterogeneous multiagent systems is solved, and efficient system control is achieved.

CN118550187BActive Publication Date: 2024-11-22XIANGJIANG LAB
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411009246.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-22
Estimated Expiration
2044-07-26

AI Technical Summary

Technical Problem

The prior art has the problem of low control efficiency in the control of agent system, especially in heterogeneous multiagent systems, where the convergence time of the algorithm depends on the control parameters, making it difficult for the system to quickly and effectively control.

Method used

By obtaining agent information, determining directed communication topology information, and using a preset distributed optimization algorithm to perform function solutions to determine the minimized global cost function solution, thereby achieving collaborative control.

Benefits of technology

It improves the accuracy and efficiency of system control, can quickly determine the solution to minimize the global cost function, and solves the problem of low system control accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118550187B_ABST
    Figure CN118550187B_ABST
Patent Text Reader

Abstract

The embodiment of the present invention provides a method and device for fully distributed optimization of a heterogeneous multi-intelligent system under a preset time, which relates to the technical field of intelligent system control technology. The method includes: obtaining agent information of agents included in the target system; based on the agent information, determining the directed communication topology information between the agents; performing function solving calculations on the directed communication topology information through a preset distributed optimization algorithm to determine a solution that minimizes the global cost function; and collaboratively controlling the target system according to the solution that minimizes the global cost function. Through the present invention, the problem of low agent control efficiency is solved, thereby achieving the effect of improving the control efficiency and accuracy of the intelligent system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of intelligent agent control technology, and more specifically, to a fully distributed optimization method and device for a heterogeneous multi-intelligent system under a preset time. Background Art

[0002] In recent years, the cooperative control problem of general linear multi-agent systems has attracted the attention of many experts and scholars, such as formation tracking problems and consistency problems. In the study of distributed optimization problems, the idea of ​​characterizing the dynamics of participating agents as general linear systems has also emerged, and relevant research results have been achieved.

[0003] Since the individuals participating in different distributed optimization tasks have different dynamic characteristics, more and more scholars have begun to consider the dynamics of the individuals involved in the study of optimization problems; although existing technologies can quickly find the minimum of the global cost function, the convergence time of the algorithm depends on some control parameters, which makes it impossible to quickly and effectively control the intelligent system.

[0004] There is currently no better solution to the above problems. Summary of the invention

[0005] The embodiments of the present invention provide a method and device for fully distributed optimization of a heterogeneous multi-intelligent system under a preset time, so as to at least solve the problem of low control efficiency of the intelligent system in the related art.

[0006] According to an embodiment of the present invention, a fully distributed optimization method for a heterogeneous multi-intelligence system under a preset time is provided, comprising:

[0007] Acquire agent information of agents included in the target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled;

[0008] Based on the agent information, determining directed communication topology information between the agents, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weight adjacency matrix between the agents;

[0009] Performing function solving calculation on the directed communication topology information by using a preset distributed optimization algorithm to determine a function solution that minimizes the global cost;

[0010] The target system is cooperatively controlled according to the minimized global cost function solution.

[0011] In an exemplary embodiment, before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution to minimize the global cost function, the method further includes:

[0012] Based on the directed communication topology information, determining whether a target condition is satisfied;

[0013] According to whether the target conditions are met, a distributed optimization function is determined, and the function solution calculation is determined based on the distributed optimization function.

[0014] In an exemplary embodiment, determining whether a target condition is satisfied based on the directed communication topology information includes:

[0015] Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph;

[0016] Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition;

[0017] Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

[0018] According to another embodiment of the present invention, a fully distributed optimization device for a heterogeneous multi-intelligence system under a preset time is provided, comprising:

[0019] An information collection module, used to obtain agent information of agents included in the target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled;

[0020] An information determination module, configured to determine directed communication topology information between the agents based on the agent information, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weighted adjacency matrix between the agents;

[0021] A function solving module, used to perform function solving calculation on the directed communication topology information by using a preset distributed optimization algorithm to determine a function solution that minimizes the global cost;

[0022] A collaborative control module is used to collaboratively control the target system according to the minimized global cost function solution.

[0023] In an exemplary embodiment, the apparatus further comprises:

[0024] A condition determination module, used to determine satisfaction of a target condition based on the directed communication topology information before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution to minimize a global cost function;

[0025] The function determination module is used to determine the distributed optimization function according to whether the target conditions are met, and the function solution calculation is determined based on the distributed optimization function.

[0026] In an exemplary embodiment, determining whether a target condition is satisfied based on the directed communication topology information includes:

[0027] Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph;

[0028] Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition;

[0029] Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

[0030] According to yet another embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps in the above method embodiment when running.

[0031] According to yet another embodiment of the present invention, there is provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in the above method embodiment.

[0032] Through the present invention, since the directed communication topology information is calculated by a global distributed algorithm, the minimized global cost function solution for system control can be quickly determined. Therefore, the problem of low system control accuracy can be solved, thereby achieving the effect of improving system control accuracy and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a hardware structure block diagram of a mobile terminal of a fully distributed optimization method for a heterogeneous multi-intelligent system under a preset time according to an embodiment of the present invention;

[0034] Figure 2 is a flowchart of a fully distributed optimization method for a heterogeneous multi-intelligence system under a preset time according to an embodiment of the present invention;

[0035] Figure 3 This is the simulation effect according to the specific embodiment of the present invention. Figure 1 ;

[0036] Figure 4 This is the simulation effect according to the specific embodiment of the present invention. Figure 2 ;

[0037] Figure 5 This is the simulation effect according to the specific embodiment of the present invention. Figure 3 ;

[0038] Figure 6 It is a structural block diagram of a fully distributed optimization device for a heterogeneous multi-intelligent system under a preset time according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings and in combination with the embodiments.

[0040] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0041] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 1 is a hardware structure block diagram of a mobile terminal of a fully distributed optimization method for heterogeneous multi-intelligence systems under a preset time according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0042] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as a computer program corresponding to a fully distributed optimization method for a heterogeneous multi-intelligence system under a preset time in an embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0043] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0044] In this embodiment, a fully distributed optimization method for a heterogeneous multi-intelligence system under a preset time is provided. Figure 2 is a flowchart of a fully distributed optimization method for a heterogeneous multi-intelligence system under a preset time according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0045] Step S201, obtaining agent information of agents included in a target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled;

[0046] In this embodiment, the intelligent agent can be an intelligent device or system that needs to complete a certain task within a limited or fixed time, such as economic scheduling in a smart grid, optimal coordination of multiple mobile robots or manipulators, source positioning in a wireless sensor network, and optimized formation of multiple drones, etc. The target system is a device or system that includes these wireless sensors, manipulators, local power grids, drones and other intelligent agents.

[0047] Assume that in the game problem there is Agents participate in decision-making, and each agent is regarded as a node in the topological graph. The set of agents composed of all nodes is recorded as If the agent and If there is information exchange, and Between the edges , use Represents the set of all edges (i.e. the aforementioned agent information).

[0048] Step S202, based on the agent information, determining directed communication topology information between the agents, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weighted adjacency matrix between the agents;

[0049] In this embodiment, it can be express A real matrix, using triples represents a directed communication topology graph between agents, where is a collection of nodes, is the edge set, It is a picture The weighted adjacency matrix represents the weighted edges between nodes; Representation diagram Agent There are directed edges between them, which represent the agents Can get the intelligent agent Related information, Cannot obtain Information about yes Neighbor, Representing an Agent The set of all neighbors for the weighted adjacency matrix when When, define (The specific value is ), otherwise ,In particular, it is stipulated that the topological graph has no self-loops, i.e. ; Define the agent The in-degree and out-degree of and , if for each agent, its out-degree is equal to its in-degree, then the topological graph is called a weighted balanced directed graph; define the directed edge sequence From the intelligent agent To Agent If there is such a directed path between any two nodes, the directed graph is called strongly connected; is the out-degree matrix of the graph, and the Laplacian matrix is According to the definition of the Laplace matrix, its row and column sums are both zero, that is, for the vector (i.e., a column vector or row vector whose elements are all 1), there is the equation Established (i.e. ), assuming the matrix The characteristic value of ,in, hour, , , moreover, a directed graph is weight-balanced if and only if ,in, for The transpose of .

[0050] The symbols in this paper follow standard mathematical definitions and can be referenced as follows:

[0051] If the following inequalities are satisfied:

[0052] ;

[0053] Then it is called is a convex function.

[0054] If a differentiable function Satisfies the following inequality:

[0055] ;

[0056] In the formula, , then it is called a function exist Above is - Strongly convex.

[0057] Step S203, performing function solving calculation on the directed communication topology information by using a preset distributed optimization algorithm to determine a function solution that minimizes the global cost;

[0058] In this embodiment, the agent information interaction network is composed of A weighted balanced strongly connected directed graph consisting of , the goal of each agent is to minimize its own cost function. In order to control the output of all agents with high-order dynamics to reach the optimal point of the optimization problem and further protect the gradient privacy information of the participating agents, a fully distributed optimization algorithm can be used to control the agents to obtain the optimal solution to the optimization problem (i.e., minimize the global cost function solution) at a specified time point to achieve optimal control.

[0059] Specifically, it is achieved through the following formula:

[0060] (Formula 1)

[0061] In the above algorithm, It is Agent Dynamics System The state variables in is the constant matrix in the system, are the corresponding number of rows and columns of the corresponding matrix, is the control input, is the control output, is an auxiliary variable, is the derivative of the corresponding variable, is the local feedback gain, is the solution of the following equation:

[0062] ;

[0063] in, . Is in the system Satisfy the conditions The parameters here are is to satisfy the equation and Here, It is an intelligent agent local demand.

[0064] Agent The closed loop form is:

[0065] (Formula 2)

[0066] In this system:

[0067] It means to express the vector as a column vector. is the derivative of the above corresponding variable, and the meaning of the corresponding variable is the same as that of the corresponding variable in Formula 1.

[0068] , yes The identity matrix of .

[0069] Step S204: collaboratively control the target system according to the minimized global cost function solution.

[0070] In this embodiment, after obtaining the minimized global cost function solution, the intelligent agent is coordinated and controlled according to the minimized global algebraic function solution, so as to optimally complete the relevant tasks of the intelligent agent in the minimum time, improve the system's control efficiency over each intelligent agent, and also reduce energy consumption; for example, for the economic dispatch problem of the power system, the coordinated operation of the power grid that can meet the power generation demand constraints and power generation capacity is achieved by minimizing the global cost function solution.

[0071] In an optional embodiment, before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution for minimizing the global cost function, the method further includes:

[0072] Step S20301, determining whether a target condition is satisfied based on the directed communication topology information;

[0073] Step S20302: determining a distributed optimization function according to whether the target condition is satisfied, and the function solving calculation is determined based on the distributed optimization function.

[0074] In this embodiment, before performing calculation, it is first determined whether the directed communication topology information satisfies relevant conditions, so as to adaptively adjust the calculation method and accurately perform corresponding processing.

[0075] In an optional embodiment, determining whether a target condition is satisfied based on the directed communication topology information includes:

[0076] Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph;

[0077] Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition;

[0078] Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

[0079] In this embodiment, several commonly used assumptions are made:

[0080] Assumption 1: Interaction topology between agents is a weighted balanced strongly connected directed graph;

[0081] Assumption 2: Agent exist The local cost function on is differentiable and -strongly convex (corresponding to the first condition above), its gradient satisfies the constant The Lipschitz condition, where and are all positive numbers (corresponding to the second condition mentioned above);

[0082] Assumption 3: For any (corresponding to the aforementioned constant matrix) is controllable, and:

[0083] ;

[0084] in, It is a system The constant matrix in , is a state variable, is the control input, is the control output, and is the corresponding dimension in the constant matrix, i.e. is the corresponding number of rows and columns of the corresponding matrix; (corresponding to the third condition mentioned above)

[0085] Lemma 1: When Assumption 3 holds, the triple is the solution of the following equation:

[0086] ;

[0087] in, ;

[0088] Assuming that the matrix in 3 satisfies the row rank condition, the solution of the equation in Lemma 1 can be obtained, which provides convenience for the following design of distributed optimization algorithm.

[0089] Lemma 2: Assume is the optimal solution for the optimization problem, then the following conclusions hold:

[0090] (Formula 3)

[0091] On the contrary, if for a feasible value If the condition (Formula 3) is met, then is the global optimum of the optimization problem. Here, It is an intelligent agent local demand.

[0092] In addition, when assumptions 1 and 3 hold, if there exists a matrix Make is a Hurwitz matrix, then the equilibrium point of the system The corresponding output is the optimal output of the optimization problem, where .

[0093] Based on the above conditions, the solution to minimize the global cost function is obtained.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0095] The following is an explanation using specific examples.

[0096] Assume that in the game problem there is Agents participate in decision-making, and each agent is regarded as a node in the topological graph. The set of agents composed of all nodes is recorded as If the agent and If there is information interaction, and Between the edges ,use Represents the set of all edges. Use triples represents a directed communication topology graph between agents, where is a collection of nodes, is the edge set, It is a picture The weighted adjacency matrix represents the weighted connections between nodes. Representation diagram Agent There are directed edges between them, which represent the agents Can get the intelligent agent Related information, Cannot obtain Information about yes Neighbor, Representing an Agent The set of all neighbors for the weighted adjacency matrix when When, define (The specific value is the edge ), otherwise In particular, it is stipulated that the topological graph has no self-loops. ; Define the agent The in-degree and out-degree of and , if for each agent, its out-degree is equal to its in-degree, then the topological graph is called a weighted balanced directed graph; define the directed edge sequence From the intelligent agent To Agent If there is such a directed path between any two nodes, the directed graph is called strongly connected. is the out-degree matrix of the graph, and the Laplacian matrix is According to the definition of the Laplace matrix, the sum of its rows and columns is zero. That is to say, for the vector , there is an equation Established, assuming that the matrix The characteristic value of ,in, hour, , , moreover, a directed graph is weight-balanced if and only if .

[0097] The symbols in this paper follow standard mathematical definitions and can be referenced as follows:

[0098] If the following inequalities are satisfied:

[0099] ;

[0100] Then it is called is a convex function;

[0101] If a differentiable function Satisfies the following inequality:

[0102] ;

[0103] In the formula, , then it is called a function exist Above is - Strongly convex.

[0104] Here is the problem description

[0105] The agent information interaction network is composed of A weighted balanced strongly connected directed graph consisting of , the goal of each agent is to minimize its own cost function and satisfy certain coupled linear equality constraints:

[0106] (Formula 4)

[0107] In the above questions, The output variables of the agents are It is represented by the cost function Description, and , is the global coupling constraint, represents the overall demand of all agents, is a constant representing the local demand of each agent.

[0108] No. The high-order dynamics system of an agent is described as:

[0109] (Formula 5)

[0110] In this dynamic system, It's the state. is the control input, is the output of the agent, and the constant matrix in the system is .

[0111] In order to control the outputs of all agents with high-order dynamics to reach the optimal point of the optimization problem, a distributed optimization algorithm is designed to obtain a solution that can minimize the global cost function.

[0112] Note 1: Compared with the existing distributed optimization problems of general linear multi-agent systems, the optimization problem of the general linear system considered takes into account the influence of global coupling constraints on the optimal solution. Therefore, the existing distributed algorithms cannot be used to solve the described optimization problem.

[0113] Here are some commonly used assumptions:

[0114] Assumption 1: Interaction topology between agents It is a weighted balanced strongly connected directed graph.

[0115] Assumption 2: Agent exist The local cost function on is differentiable and - is strongly convex, and its gradient satisfies the constant The Lipschitz condition, where and They are all normal numbers.

[0116] Assumption 3: For any is controllable and:

[0117] ;

[0118] Lemma 1: When Assumption 3 holds, the triple is the solution of the following equation:

[0119] ;

[0120] in, .

[0121] Assuming that the matrix in 3 satisfies the row rank condition, the solution of the equation in Lemma 1 can be obtained, which provides convenience for the following design of distributed optimization algorithm.

[0122] Lemma 2: Assume is the optimal solution for the optimization problem, then the following conclusions hold:

[0123] (Formula 6)

[0124] On the contrary, if for a feasible value If the conditions are met, is the global optimal solution to the optimization problem.

[0125] The following are the main conclusions

[0126] Distributed Algorithm Design

[0127] In order to further protect the gradient privacy information of participating agents, a fully distributed optimization algorithm will be designed to control the agents to obtain the optimal solution to the optimization problem at a specified time point:

[0128] (Formula 7)

[0129] In the above algorithm, is an auxiliary variable.

[0130] Agent The closed loop form is:

[0131] (Formula 8)

[0132] In this system:

[0133] ,

[0134] .

[0135] The following is the stability analysis

[0136] Theorem 1: When Assumptions 1 and 3 hold, if there exists a matrix Make is a Hurwitz matrix, then the equilibrium point of the system The corresponding output It is the optimal output of the optimization problem.

[0137] The following are the simulation results and analysis

[0138] In order to verify the correctness and effectiveness of the distributed algorithm, such as Figure 3-Figure 5 As shown, the dynamics of the intelligent agent is selected as a linear reduced-order aircraft system, and:

[0139] ,

[0140] in, Respectively represent the sideslip angle, rolling angular velocity, and yaw angular rate, They represent the aileron deflection, differential tail deflection, roll thrust vector deflection, rudder deflection, and yaw thrust deflection of an aircraft. When the lift is 10,000 ft, the parameter matrix of the selected system is:

[0141] , .

[0142] This simulation example considers the distributed optimization problem of a heterogeneous multi-agent system with a directed communication network consisting of 6 agents. The communication topology diagram is .

[0143] The cost function of all agents is:

[0144] ;

[0145] here, ;

[0146] The calculated parameter matrix is ​​chosen as:

[0147] ;

[0148] By comparison Figure 3 and Figure 4 ,in, Figure 3 The time base generator function (TBG function) is not introduced. Figure 4 When the time base generator function is introduced into the algorithm, it can be seen that when the time base generator function is introduced into the algorithm, the algorithm can effectively control the agent to reach the global optimal solution at the preset time, and when the preset time is 0.6s, Figure 5 It can be seen that the system can still control the participating individuals to reach the optimal solution of the optimization problem in a shorter time. Figure 3 , 4 and 5, all solid lines from top to bottom represent the optimal solutions reached by agents 3, 5, 6, 4, 2, and 1 respectively.

[0149] In this embodiment, a fully distributed optimization device for a heterogeneous multi-intelligent system under a preset time is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0150] Figure 6 is a structural block diagram of a fully distributed optimization device for a heterogeneous multi-intelligent system under a preset time according to an embodiment of the present invention, such as Figure 6 As shown, the device comprises:

[0151] An information collection module 61 is used to obtain agent information of agents included in a target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled;

[0152] An information determination module 62, configured to determine directed communication topology information between the agents based on the agent information, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weighted adjacency matrix between the agents;

[0153] A function solving module 63, configured to perform function solving calculations on the directed communication topology information by using a preset distributed optimization algorithm to determine a function solution that minimizes the global cost;

[0154] The collaborative control module 64 is used to collaboratively control the target system according to the minimized global cost function solution.

[0155] In an optional embodiment, the device further comprises:

[0156] A condition determination module, used to determine satisfaction of a target condition based on the directed communication topology information before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution to minimize a global cost function;

[0157] The function determination module is used to determine the distributed optimization function according to whether the target conditions are met, and the function solution calculation is determined based on the distributed optimization function.

[0158] In an optional embodiment, determining whether a target condition is satisfied based on the directed communication topology information includes:

[0159] Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph;

[0160] Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition;

[0161] Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

[0162] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0163] An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps in the above method embodiment when running.

[0164] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0165] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in the above method embodiment.

[0166] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0167] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0168] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A fully distributed optimization method for heterogeneous multi-intelligence systems under a preset time, characterized in that: include: Acquire agent information of agents included in the target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled; Based on the agent information, determining directed communication topology information between the agents, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weight adjacency matrix between the agents; The directed communication topology information is subjected to function solving calculations by a preset distributed optimization algorithm to determine a solution that minimizes the global cost function; wherein the preset distributed optimization algorithm is implemented by the following formula: In the formula, It is Agent Dynamics System The state variables in is the constant matrix in the system, are the corresponding number of rows and columns of the corresponding matrix, is the control input, is the control output, is an auxiliary variable, is the derivative of the corresponding variable, is the local feedback gain, is the solution of the following equation: in, ; Is in the system Satisfy the conditions Parameters, is to satisfy the equation and The time base generator function, It is an intelligent agent local needs; The target system is cooperatively controlled according to the minimized global cost function solution.

2. The method according to claim 1, characterized in that Before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution to minimize the global cost function, the method further includes: Based on the directed communication topology information, determining whether a target condition is satisfied; According to whether the target conditions are met, a distributed optimization function is determined, and the function solution calculation is determined based on the distributed optimization function.

3. The method according to claim 2, characterized in that The determining, based on the directed communication topology information, whether a target condition is satisfied includes: Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph; Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition; Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

4. A fully distributed optimization device for heterogeneous multi-intelligence systems under preset time, characterized in that: include: An information collection module, used to obtain agent information of agents included in the target system, wherein the agent information includes the number of agents, whether they participate in decision-making, and information interaction between agents, and the target system includes a system that needs to be collaboratively controlled; An information determination module, configured to determine directed communication topology information between the agents based on the agent information, wherein the directed communication topology information is represented by a triple, and the triple includes a node set, an edge set, and a weighted adjacency matrix between the agents; A function solving module is used to perform function solving calculation on the directed communication topology information through a preset distributed optimization algorithm to determine a function solution that minimizes the global cost; wherein the preset distributed optimization algorithm is implemented by the following formula: In the formula, It is Agent Dynamics System The state variables in is the constant matrix in the system, are the corresponding number of rows and columns of the corresponding matrix, is the control input, is the control output, is an auxiliary variable, is the derivative of the corresponding variable, is the local feedback gain, is the solution of the following equation: in, ; Is in the system Satisfy the conditions Parameters, is to satisfy the equation and The time base generator function, It is an intelligent agent local needs; A collaborative control module is used to collaboratively control the target system according to the minimized global cost function solution.

5. The device according to claim 4, characterized in that The device also includes: A condition determination module, used to determine satisfaction of a target condition based on the directed communication topology information before performing function solving calculation on the directed communication topology information by a preset distributed optimization algorithm to determine a solution to minimize a global cost function; The function determination module is used to determine the distributed optimization function according to whether the target conditions are met, and the function solution calculation is determined based on the distributed optimization function.

6. The device according to claim 5, characterized in that The determining, based on the directed communication topology information, whether a target condition is satisfied includes: Based on the directed communication topology information, determining whether the interaction topology graph between the agents is a weighted balanced strongly connected directed graph; Or, based on the directed communication topology information, determining whether the local cost function of the agent satisfies a preset first condition, and whether the gradient of the local cost function of the agent satisfies a preset second condition; Or, based on the directed communication topology information, determine whether the constant matrix of any of the agents is controllable and whether the constant matrix satisfies a preset third condition.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 3 when executed.

8. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the method according to any one of claims 1 to 3.