A multi-agent cooperative control formation method

Abstract

The invention provides a multi-agent collaborative regulation and control formation method, which comprises the following steps that: after formation initialization is completed by a multi-agent system, carrying out formation, and information interaction among nodes is completed; enabling each following node to demodulat neighbor node time slot table information in two hops from the received information; finishing network maintenance of the distributed network, demodulating third-order state information of neighbor nodes in two hops, finishing correction of second-order information of time delay state information based on a prediction mechanism, taking a prediction result as a real-time state of the nodes, and finishing updating of the state information of the neighbor nodes in the two hops; Inputting the updated state information of the neighbor nodes in the two hops into a consistency regulation and control mechanism, completing updating of node control output based on the state information of the neighbor nodes in the two hops, and inputting a control output result into a power system of the following node to serve as the acceleration of the following node; And circularly executing the formation process until the task is finished. The method can effectively resist the information interaction delay constraint in the network environment, and guarantees the formation realization of the unmanned system.

Description

technical field [0001] The invention belongs to the technical field of networked control, and more specifically, relates to a multi-agent coordinated control formation method. Background technique [0002] The theory of multi-agent cooperative control is the basis of multi-agent swarm system formation control, and its main task is the movement coordination of agent nodes. In the field of control mechanism based on consistency theory, the agent node interacts with the nodes in the system, and updates the control output of the node itself based on the interaction result and the expected formation goal, thereby realizing the desired formation. The realization of desired formation by multi-agents is the basis for performing subsequent tasks. It can be said that the performance of multi-agent cooperative control technology directly affects the overall performance of the entire system. [0003] As a substitute for human combat, the air-ground integrated unmanned formation mainly...

AI Technical Summary

Problems solved by technology

[0016] A reasonable model has not been established for the analysis of the control mechanism oriented to network constraints and the optimization theory based on the prediction mechanism. There are unreasonable assumptions about the network delay parameters in the control academic circle, such as the belief that the network delay is cyclical, and the network delay is greater than an interaction cycle. At the same time, the relationship between network delay parameters and control system parameters has not been established, and network delay is related to parameters such as system topology and system traffic. The establishment and introduction of network delay models based on control system parameters are very important.

[0017] Furthermore, the control mechanism only considers the influence of network parameter constraints on the control system, and the specific form of the control mechanism based on the distributed network model has not yet been established.

Images