Multi-agent formation collaboration device and control method thereof

Abstract

The invention discloses a multi-agent formation collaboration device and a control method thereof, belongs to the field of multi-agent collaboration, and provides a multi-agent formation collaborationdevice and a control method thereof. Folding airbags are installed at the edges of agents, and when the distance between every two adjacent agents is too short, a control module controls the foldingairbags to be inflated and stretched; when the agents move, the self-protruding adsorption balls collide with the inner wall of the unfolded folding airbag and are attached to the inner wall of the folding airbag, so the strength and hardness of the surfaces of the self-protruding adsorption balls and the inner wall of the folding air bag can be effectively improved, resilience force generated after the folding airbag is collided can be reduced; and the two colliding intelligent bodies are not prone to generating large displacement due to counter-acting force of collision force; therefore, while the collision stability of the agents is improved, when the folding airbags bear the collision force, the folding airbags attract one another to form a whole, the maintaining effect on the stability of the agents is better, so the damage to the multi-agent when the collaborative data is modified is effectively reduced.

Description

technical field [0001] The present invention relates to the field of multi-agent cooperation, and more specifically, relates to a multi-agent formation cooperation device and a control method thereof. Background technique [0002] Multi-agent system (MAS) is a new distributed computing technology. Since its appearance in the 1970s, it has developed rapidly and has become a thinking method and tool for complex system analysis and simulation. [0003] A multi-agent system (M.A.S.) is a computing system composed of multiple agents interacting in an environment. Multi-agent systems can also be used to solve problems that are difficult for separate agents and single-layer systems. Intelligence can be realized by some method, function, process, search algorithm or reinforcement learning. Despite considerable overlap, however, a multi-agent system does not always behave consistently with an agent-based model (ABM). The goal of ABM is to find explanations for the collective beha...

AI Technical Summary

Problems solved by technology

[0004] A multi-agent collaborative system can usually cooperate with UAV formations, unmanned vehicle formations, and unmanned ship formations, etc., but for different formations, the system's collaborative control parameters have certain changes, resulting in the replacement of objects in the agent. Finally, the coupling strength of the formation will change. In order to ensure a suitable coupling strength, the collaborative data usually needs to be tested, and the collaborative data is changed in the continuous test. During the test, it is easy for the distance between two adjacent agents to exceed If the distance is too large or the distance is too small, for the agents with too small distance, it is very easy to collide with each other, resulting in certain damage to the agent.

Images