Distributed cooperative security control method applied to multi-agent system

Abstract

The invention belongs to the technical field of agent safety control, and relates to a distributed cooperative safety control method applied to a multi-agent system, and the method comprises the steps: S1, building a kinematic model of an omnidirectional mobile agent, and generating a nominal speed instruction according to a preset track; s2, determining position information of each agent and broadcasting the position information to other agents; s3, determining a candidate zero-crossing obstacle function; s4, generating a security constraint by using the candidate zero-crossing obstacle function; s5, determining a loss function based on the nominal speed, and solving a control quantity which enables the loss function to be minimum on the premise of meeting security constraints as a security control instruction; and S6, controlling the intelligent body to move according to the safety control instruction. The method has the beneficial effects that in a large-scale agent cluster, even if global information and a central node are lacked, collision among the agents can be effectively avoided, meanwhile, falling into a local optimal solution is avoided, and it is ensured that the agents are safely navigated to a specified target point.

Description

Technical Field

[0001] The present invention relates to the field of intelligent agent safety control technology, and in particular to a distributed collaborative safety control method applied to a multi-agent system. Background Art

[0002] Safety means that in a multi-agent swarm composed of multiple robots, any two agents are kept as far apart as possible. The goal of safe control is to modify the nominal control instructions as little as possible without violating safety constraints, so that each agent can safely navigate to the specified target point.

[0003] With the development of artificial intelligence and robotics, multi-agent systems (MAS) have shown their unique advantages in many fields, including formation flying, cluster movement, multi-robot collaborative operations, etc. These systems are composed of multiple agents with partial perception, communication, computing and execution capabilities. They are interconnected through the network and show a high degree of autonomy, fault tolerance, coordination and flexibility. However, when the paths of multiple agents intersect, how to avoid collisions between agents becomes a technical challenge.

[0004] After decades of development, many safety control methods have been proposed. They mainly include global motion planning based on path planning and trajectory optimization, and local motion planning based on reinforcement learning. The path planning method can plan a safe path based on environmental information to avoid deadlock; however, it relies on accurate perception of the environment, requires a lot of computing resources, and is prone to fall into local optimal solutions. Reinforcement learning methods can adapt to complex obstacle avoidance scenarios and have the possibility of achieving optimal solutions; the disadvantage is that it requires a lot of computing resources and has limited generalization capabilities. The control obstacle function is a relatively powerful method that has been recently developed. As a controller expressed in an optimized form, it ensures system safety. The controller form is simple, the calculation is simple, and the real-time performance is strong. In recent years, with the widespread application of large-scale robot clusters, related research hotspots have begun to focus on the safety control of multi-agents. For example, self-driving cars need to avoid other vehicles and pedestrians in complex traffic environments; in cooperative operations with drones, and when drones are used to deliver express, collisions need to be avoided when the paths of multiple drones intersect. In logistics warehouses, multiple handling robots need to move goods efficiently in a space, and so on.

[0005] The key issues to be considered in multi-agent safety control include: each agent only has information about its own tasks, but lacks complete knowledge of other vehicles’ tasks and states; there is coupling between agents, that is, whether an agent will collide depends not only on its own decision, but also on the decisions of adjacent agents. The safety control of multi-agents mainly adopts centralized control algorithms and decentralized control algorithms to solve these two key issues.

[0006] Among them, the centralized control algorithm assumes that there is a central node that can communicate with all agents, and each agent uploads its own status and task information to the central node. The central node centrally calculates the control instructions of all agents and sends them to all agents. The advantage is that the central node can obtain global information and make global optimal decisions. However, considering a large-scale cluster scenario in a large space, there is no central node that can cover all agents due to the limited communication range. At the same time, each agent must upload its own status and task information to the central node. The larger the cluster, the higher the network bandwidth requirement and the slower the calculation speed. Decentralized control uses the method of pre-assigning coupling constraints to decouple. An independent safety controller is deployed on each agent, and this safety controller only calculates its own safety control instructions, so it does not need the task information of other agents. The decentralized algorithm has the advantages of fast calculation speed and easy expansion. However, since the safety controller makes decisions based on local information, it is easy to fall into the local optimal solution. Summary of the invention

[0007] Technical issues to be solved: In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a distributed collaborative safety control method applied to a multi-agent system, which solves the technical problem of how to avoid collisions between agents while avoiding falling into local optimal solutions.

[0008] Technical solution: In order to achieve the above object, the main technical solutions adopted by the present invention include: In a first aspect, the present invention provides a distributed collaborative safety control method applied to a multi-agent system, comprising: S1, establish the kinematic model of the omnidirectional mobile agent and generate the nominal speed command according to the preset trajectory; S2, determine the location information of each agent and broadcast it to other agents within the preset range of each agent; S3, determining a candidate zero-crossing barrier function; S4, generating safety constraints using candidate zero-crossing barrier functions based on the kinematic model and the relative positions of each agent and all neighboring agents within its preset range; S5, determining the loss function based on the nominal speed, and solving the control quantity that minimizes the loss function under the premise of satisfying the safety constraint as the safety control instruction; S6, controlling the movement of the intelligent body according to the safety control instructions; S7, update the position of the intelligent agent and determine whether all intelligent agents have reached the end point. If not, execute S2; if so, end.

[0009] Optionally, the kinematic model is ; in, is the position of the agent, represented by a two-dimensional column vector in the two-dimensional world coordinate system represents the position of the ith agent, It is The first derivative of the position of each agent; is the linear velocity of the agent and also the control input of the system. In the two-dimensional coordinate system, it is represented by a two-dimensional column vector Indicates The speed of an agent.

[0010] Optionally, the zero-crossing barrier function for: ; The constant , is the expected safe distance between two agents, is an additional safety margin to ensure robustness. , is the total number of agents in the cluster, symbol represents the two-norm of a vector, represents the position of the jth agent.

[0011] Optionally, a safety constraint is generated based on the candidate zero-crossing barrier function, including: Based on the zero-crossing barrier function, the safety constraint g is derived from time t as follows: ; in, is the decay rate of the barrier function, which is a constant greater than 0; is the sampling time, u is the control input, .

[0012] Optionally, the loss function is: ; in, is the nominal control quantity, i An agent provides a reference value of speed.

[0013] In a second aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of the first aspects above.

[0014] In a third aspect, the present invention provides a storage device comprising a storage medium and a processor, wherein the storage medium stores a computer program, and when the program is executed by the processor, it implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of the first aspects above.

[0015] Beneficial effects: The beneficial effects of the present invention are as follows: Compared with the prior art, the distributed collaborative safety control method for a multi-agent system of the present invention has the following significant advantages: (1) The method proposed in this paper uses a distributed optimization algorithm to enable each agent to consider the decision information of adjacent agents and achieve collaborative obstacle avoidance. This not only avoids collisions between agents, but also effectively solves the problem of local optimal solutions caused by the decentralized safety controller being based only on local information, ensuring the optimality of global path planning.

[0016] (2) The present invention distributes the safety controller to each intelligent agent. Each intelligent agent only needs to communicate with adjacent intelligent agents through the local area network, which solves the problem of limited communication range. It also improves the problem of insufficient network bandwidth caused by collecting all intelligent agent information and slow computing speed caused by solving all robot safety constraints in the cluster when facing the safety control of large-scale robot clusters.

[0017] (3) The agent broadcasts itself in the future The position information after seconds is used to share the motion intention with the adjacent agents in advance, reducing the position error caused by the movement of the agent during the calculation process. At the same time, an additional safety margin is introduced to compensate for the position prediction error and calculation error caused by network delay. This mechanism not only solves the problem of inconsistent clocks caused by network delay in multi-agent systems, but also ensures that even in the case of non-continuous calculations, the collaborative obstacle avoidance between agents can be carried out safely and effectively, thereby significantly improving the robustness of the multi-agent system in avoiding obstacles in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A control system block diagram provided by an embodiment of the present invention; Figure 2 A schematic diagram of the initial state of a simulation provided by an embodiment of the present invention; Figure 3 A schematic diagram of simulation results of a distributed safety controller provided by an embodiment of the present invention; Figure 4 A schematic diagram of simulation results of a distributed safety controller provided by an embodiment of the present invention; Figure 5 A schematic diagram of simulation results of a centralized safety controller provided by an embodiment of the present invention; Figure 6 A physical experimental platform for the feasibility of the embodiments of the present invention; Figure 7 A diagram of physical experiment results provided by an embodiment of the present invention; Figure 8A flow chart of a distributed collaborative safety control method applied to a multi-agent system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0020] The present invention is aimed at the safety requirements of omnidirectional mobile robots performing tasks in large-scale robot cluster scenarios, and designs a distributed collaborative safety control method applied to multi-agent systems. On the one hand, any agent in the cluster only knows its own control target; on the other hand, the agent can only communicate with adjacent agents, and there is no central node that can obtain global information. The goal of this method is that the agent avoids collisions with other agents and reaches the target position based only on the information of itself and adjacent agents. The key feature of this method is that during the real-time interaction between agents, it can always ensure that the states of any two agents in the cluster are within the safety set. And under the premise that the agent does not have global information, by communicating with adjacent agents, the nominal control instructions are modified as little as possible to complete their respective task goals.

[0021] This method formulates the safety control problem as a quadratic programming problem with linear inequality constraints by constructing a robust control obstacle function and generating safety constraints, and introduces a distributed optimization algorithm to solve it. In this invention, the present invention makes several important improvements to address the lack of global information and coupling constraints between agents in a multi-agent system. First, by designing candidate zero-crossing obstacle functions and constructing safety constraints, it is ensured that the agent is safe for each adjacent agent; secondly, the safety control problem is converted into a quadratic programming problem, and the control input is optimized to meet all safety constraints at the same time; finally, a distributed optimization algorithm is introduced to achieve distributed iterative solution of the quadratic programming problem on each agent, avoiding the problems of limited communication range of the central node and insufficient network bandwidth. These three improvements enable the effective avoidance of collisions between agents in a large-scale agent cluster, even in the absence of global information and central nodes, and ensure that the agent safely navigates to the designated target point.

[0022] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0023] The controlled object considered in this embodiment is A multi-agent system consisting of agents with omnidirectional motion capabilities is described here using an unmanned vehicle equipped with Mecanum wheels that can perform omnidirectional motion. The kinematic model of the agent can be expressed by a first-order integrator model: ; in, is the position of the agent, represented by a two-dimensional column vector in the two-dimensional world coordinate system represents the position of the ith agent, It is The first derivative of the position of each agent; is the linear velocity of the agent and also the control input of the system. In the two-dimensional coordinate system, it is represented by a two-dimensional column vector Indicates The speed of an agent.

[0024] In addition, a path tracing algorithm is used to derive a bounded, unverified, nominal speed. The distributed safety controller designed in this embodiment is used to perform safety filtering on the nominal speed, and the output safety control instructions are given to the motion controller, which ultimately drives the intelligent body to move so that it reaches the target point without colliding with other intelligent bodies.

[0025] exist Figure 1 The overall algorithm framework of the method provided in this embodiment is shown in FIG. Figure 2 , Figure 3 , Figure 4 and Figure 5 The comparison of Matlab simulation results of the controller designed based on the algorithm of this embodiment and the distributed algorithm is demonstrated. Figure 6 The physical experiment system of the algorithm is demonstrated. The physical experiment system includes the motion capture system Optitrack and several omnidirectional mobile robots equipped with Raspberry Pi. All devices are connected to the same local area network through wireless network cards or cables, and the ROS distributed communication framework is used to realize information transmission between devices. Figure 7 Physical experimental results of the algorithm are shown.

[0026] First, refer to Figure 8 This embodiment provides a distributed collaborative safety control method applied to a multi-agent system, including: S1, establish the kinematic model of the omnidirectional mobile agent and generate the nominal speed command according to the preset trajectory.

[0027] S2, determine the location information of each agent and broadcast it to other agents within the preset range of each agent.

[0028] Assume that the agents in the cluster have the same control period , Greater than the time required for the distributed intelligent agent to solve the quadratic programming problem. At this moment, the agent samples its own position information through the motion capture system , and based on its current speed, predict Then the location information of the ,in Is within the current control cycle , Agent The amount of control actually performed. The agent latches this information during the control cycle And inform the neighboring agents through broadcasting. At the same time, obtain the sampling position information broadcast by the neighboring agents .

[0029] S3, determining a candidate zero-crossing barrier function.

[0030] ; in, , .

[0031] is the expected safe distance between two agents, usually the sum of the expansion circle radii of agent i and agent j.

[0032] It is an additional safety margin that can compensate for the impact of predicted position error and calculation error on system safety. The definition of prediction error is as follows The prediction error can be regarded as the measurement error caused by the sensor, and the calculation error caused by the finite-step iteration can be regarded as the actuator error. Existing papers in related fields have proved that there are sampling systems with measurement errors and actuator errors. By adding a safety margin, it can be ensured that the application of safety control quantity can still ensure that the state quantity is within the safety set. Therefore, this patent adds a safety margin term , in order to compensate for the prediction deviation and calculation error of the agent's own position information, generate control input, and still ensure that the system state strictly satisfies forward invariance.

[0033] S4, based on the kinematic model and the relative position of each agent to all its neighboring agents within a preset range, generates safety constraints using candidate zero-crossing barrier functions.

[0034] Design candidate zero-crossing barrier functions based on the relative distance between two different agents , and construct safety constraints to ensure that the agent is safe with respect to every neighboring agent.

[0035] Based on the theory of control obstacle function, About Time Take the derivative and get a safety constraint.

[0036] This constraint defines a set of control strategies that meet safety conditions by limiting the control input of the agent to ensure that there is no collision between agents. Usually, in order to ensure that there is no collision between any two agents in a multi-agent system, it is necessary to impose the same form of safety constraints on any pairwise combination of all agents.

[0037] .

[0038] Therefore, the number of constraints will increase dramatically, the speed of solving the optimization problem will slow down, and the real-time performance of the algorithm will deteriorate. In addition, building security constraints between all agents requires global agent location information, which is probably not feasible considering the actual communication range and bandwidth limitations. In view of the above problems, the present invention provides a solution in S5.

[0039] S5, determining the loss function based on the nominal speed, and solving the control quantity that minimizes the loss function under the premise of satisfying the safety constraints as the safety control instruction.

[0040] The control goal of the multi-agent swarm safety control problem is to minimize the modification of the nominal control instructions while ensuring that no collision occurs between any two agents in the swarm. Therefore, a quadratic cost function can be used to compensate for the deviation between the control quantity and the nominal control quantity, and the safety constraint can be used as a hard constraint of the optimization problem, thereby formulating the safety control problem as a quadratic programming problem with multiple linear inequality constraints. This is a convex optimization problem.

[0041] Considering that the safety control problem is a local problem, it means that when the agents are far away from each other, their dynamic behaviors may not be of concern. Therefore, if the safety constraints between agents that are far enough away are ignored, the above safety controller is still valid. At the same time, considering that the solution time of the multi-agent safety controller cannot be ignored, the controller can be modified as follows: ; ; .

[0042] Indicates that at the sampling time , a set of other agents in the agent's neighborhood, defined as follows: ; in, is the upper bound of the agent’s speed, is a constant greater than 0 defined in step 2.

[0043] It is not difficult to see that the original quadratic programming problem has coupled linear constraints. The decision of the agent depends on its neighboring agents This creates an algebraic loop that prevents the controller from being directly deployed in a distributed manner.

[0044] The coefficient matrix of the quadratic term of the original quadratic programming problem is positive definite, and the constraints are all linear inequalities. Therefore, the original quadratic programming problem is a convex optimization problem. According to the strong duality principle, we can solve the dual problem of the original quadratic programming problem: ; ; in It is a constraint The corresponding Lagrange multiplier is, , is the Lagrangian function of the original quadratic programming problem, defined as follows: .

[0045] The Lagrangian function contains two sets of variables, is the control input, is a constraint variable. Solve for the inner control input , and then the outer constraint variables are updated using the gradient ascent method.

[0046] The specific method for solving the dual problem is given as follows: 1. Initialize the initial value of the constraint variable , solve for the inner control input ; .

[0047] 2. Use the gradient ascent method to update the outer constraint variables ; .

[0048] in depending on Moment, Agent and The number of other agents in their respective neighborhoods and , defined as follows: .

[0049] 3. Order ,in is the maximum number of iterations limited by network communication delay.

[0050] Agent Update control input locally and the constraint variables associated with them .

[0051] Agent Iterates locally and updates the constraint variables related to itself When , only the local iteration result of the previous iteration round is needed and the result of iterating on neighboring agents . While calculating the local control input Only local constraint variables are required . Therefore, the method for solving the dual problem is completely distributed.

[0052] It should be noted that and The same constraint corresponds to and Agent But since they have the same iteration formula, .

[0053] S6, controlling the movement of the intelligent body according to the safety control instructions.

[0054] S7, update the position of the intelligent agent and determine whether all intelligent agents have reached the end point. If not, execute S2; if so, end.

[0055] Optionally, the kinematic model is: ; in, is the position of the agent, represented by a two-dimensional column vector in the two-dimensional world coordinate system represents the position of the ith agent, It is The first derivative of the position of each agent; is the linear velocity of the agent and also the control input of the system. In the two-dimensional coordinate system, it is represented by a two-dimensional column vector Indicates The speed of an agent.

[0056] Optionally, the zero-crossing barrier function h is: ; The constant , is the expected safe distance between two agents, is an additional safety margin to ensure robustness. , is the total number of agents in the cluster, symbol represents the two-norm of a vector, represents the position of the jth agent.

[0057] Optionally, a safety constraint is generated based on the candidate zero-crossing barrier function, including: Based on the time derivative of the zero-crossing barrier function, the safety constraint g is as follows: ; in, is the decay rate of the barrier function, which is a constant greater than 0; is the sampling time, For control input.

[0058] Optionally, the loss function is: ; in, is the nominal control quantity, i An agent provides a reference value of speed, Representative i The final position of the agent, is the actual output control quantity.

[0059] Optionally, a distributed optimization algorithm is used to solve the loss function to obtain a control input, including: Inner control input based on Lagrangian function to solve loss function , use the gradient ascent method to update the outer constraint variable λ.

[0060] The distributed safety control method designed by the present invention has the following characteristics: Unlike traditional safety control methods, the distributed safety controller does not rely on its own sensors to sense the location information of neighboring agents. Instead, it informs neighboring agents of its own location information through broadcasting, and receives the location information broadcast by neighboring agents at the same time. This method ensures that different agents can jointly handle the same problem even if their clocks are not synchronized.

[0061] By deploying the above distributed controller on each agent, each agent uses local information to solve its own safety control instructions, ensuring the safety of the multi-agent system. In addition, the safety control decision made by the control method based on the local information of each agent is globally optimal.

[0062] The input of the control method is the information of adjacent intelligent agents. Even when facing a large-scale cluster of intelligent agents in a wide range of scenarios, a single intelligent agent only needs the information of adjacent intelligent agents, which greatly reduces the network bandwidth requirement and reduces the calculation speed. It is also suitable for scenarios where no central node exists.

[0063] It should be noted that in the method proposed in the present invention, when the intelligent agent calculates the local security problem in its own neighborhood, it is completely based on local information and there is no need to obtain the global position information of all intelligent agents.

[0064] Specifically, each agent Broadcast its own location information and receive broadcast information from neighboring agents , can only obtain the location of neighboring agents with which it communicates directly. Neighborhood Set The definition of depends on the relative distance between agents, and the construction of the neighborhood set depends only on the location information of local agents, does not require the knowledge of all agents, and is completely implemented locally in the agent.

[0065] The present invention uses a distributed optimization algorithm to ensure that when the agent solves the local optimization problem locally, it is only based on the decision information iteration of the agent in the neighborhood, rather than the decision information of all agents in the cluster, thereby ensuring the fully distributed characteristics of the safety control method proposed by the present invention.

[0066] This processing flow is consistent with the core feature of the present invention, namely, each intelligent agent completes security control only through local interaction, avoiding centralized global information dependence and solving the problems of limited communication range and insufficient network bandwidth.

[0067] The present invention selects the obstacle avoidance problem of 8 omnidirectional Mecanum wheeled vehicles whose kinematic models are approximately first-order integrators and avoid each other to reach their respective target points in a venue as another embodiment, and combines Figure 2 , Figure 3 , Figure 4 and Figure 5 Details are as follows: In the simulation environment, eight circular agents with a radius of 0.15 m are considered. , initial moment The initial position of the agent And mission target point The configuration is shown in Table 1.

[0068] Table 1

[0069] Define the agent's nominal controller .use Update the agent's state, is the simulation step size.

[0070] The controller parameters of the distributed safety control simulation experiment are shown in Table 2 below.

[0071] Table 2

[0072] Under the same initial conditions, the distributed safety controller proposed in this invention and the existing centralized safety controller and decentralized safety controller are respectively applied in the multi-agent system. The motion trajectory of each agent is as follows: Figure 2 , Figure 3 , Figure 4 and Figure 5 The simulation results show that , any two agents in the cluster and The relative distances are In the cluster, the minimum agent spacing distances between the distributed safety controller, the centralized safety controller and the distributed safety controller of this embodiment are , and Although the distributed safety controller has a slightly lower separation distance than the centralized solution due to sampling errors and iterative calculation errors, its actual safety distance still strictly meets the system requirements by adding a safety margin, thus ensuring the safety of the multi-agent system. This shows that the designed distributed controller can effectively coordinate local constraints while avoiding reliance on global information, achieving safety performance similar to that of the centralized solution.

[0073] Under the same simulation time and parameter settings, the agent clusters of the distributed safety controller, the centralized safety controller and the distributed safety controller designed by the present invention respectively spend ,and and This is because the security control decisions made by the distributed security controller and the centralized security controller are all globally optimal, while the agents deployed with the distributed security controller make local optimal decisions based on their own local information, so the output security control instructions are more conservative and take longer to reach the target point.

[0074] In order to further verify the feasibility of the algorithm designed in this invention, Figure 6 Physical experiments were carried out on the physical experimental platform shown in the figure. The experimental results are shown in Figure 7 shown.

[0075] The distributed safety controller proposed in the present invention is deployed on the Raspberry Pi development board of each intelligent agent. Each intelligent agent constructs a local safety problem locally based on its own and adjacent intelligent agent's position information and its own nominal speed, and uses the Lagrangian relaxation method to iteratively solve it locally. Iteration, Agent use Iteration results And the iteration results of adjacent agents Calculate the Lagrange multiplier for the current iteration , and use Update Round iteration results , and then Notify neighboring agents through broadcasting. The result of the iteration As the actual speed instruction of the agent at the current moment .

[0076] Finally, the safety control instructions output by the safety controller (deployed on the Raspberry Pi) are sent to the motion controller (deployed on the underlying development board) through the serial port to drive the intelligent body to move. Through this process, the position of the intelligent body is changed, completing the safety control at the current moment.

[0077] The results of physical experiments show that the distributed safety controller of this embodiment can navigate along the preset path from the starting point, and when the preset paths of multiple agents conflict, the agents coordinate with each other to avoid obstacles and finally reach their respective target locations safely. The safety decision made is globally optimal, avoiding the deadlock problem of the safety controller when the paths of multiple agents conflict.

[0078] In a second aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of the first aspects above.

[0079] In a third aspect, an embodiment of the present invention provides a storage device, comprising a storage medium and a processor, wherein the storage medium stores a computer program, and when the program is executed by the processor, it implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of the first aspects above.

[0080] It should be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0081] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include these modifications and variations.

[0082] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A distributed collaborative safety control method for a multi-agent system, characterized in that: include: S1, establish the kinematic model of the omnidirectional mobile agent and generate the nominal speed command according to the preset trajectory; S2, determine the location information of each agent and broadcast it to other agents within the preset range of each agent; S3, determining a candidate zero-crossing barrier function; S4, generating safety constraints using candidate zero-crossing barrier functions based on the kinematic model and the relative positions of each agent and all neighboring agents within its preset range; S5, determining the loss function based on the nominal speed, and solving the control quantity that minimizes the loss function under the premise of satisfying the safety constraint as the safety control instruction; S6, controlling the movement of the intelligent body according to the safety control instructions; S7, update the position of the intelligent agent and determine whether all intelligent agents have reached the end point. If not, execute S2; if so, end.

2. A distributed collaborative safety control method for a multi-agent system according to claim 1, characterized in that: The kinematic model is: ; in, is the position of the agent, represented by a two-dimensional column vector in the two-dimensional world coordinate system represents the position of the ith agent, It is The first derivative of the position of each agent; is the linear velocity of the agent and also the control input of the system. In the two-dimensional coordinate system, it is represented by a two-dimensional column vector Indicates The speed of an agent.

3. A distributed collaborative safety control method for a multi-agent system according to claim 2, characterized in that: The zero-crossing barrier function for: ； The constant , is the expected safe distance between two agents, is an additional safety margin to ensure robustness. , is the total number of agents in the cluster, symbol represents the two-norm of a vector, represents the position of the jth agent.

4. A distributed collaborative safety control method for a multi-agent system according to claim 3, characterized in that: Generate safety constraints based on candidate zero-crossing barrier functions, including: Based on the zero-crossing barrier function, the safety constraint g is derived from time t as follows: ； in, is the decay rate of the barrier function, which is a constant greater than 0; is the sampling time, is the control input, .

5. A distributed collaborative safety control method for a multi-agent system according to claim 4, characterized in that: The loss function is: ； in, is the nominal control quantity, i An agent provides a reference value of speed.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of claims 1 to 5.

7. A storage device comprising a storage medium and a processor, wherein the storage medium stores a computer program, characterized in that: When the processor executes the computer program, it implements a distributed collaborative safety control method applied to a multi-agent system as described in any one of claims 1 to 5.

Images