Multi-agent System Fault Detection Method Based on Comprehensive Judgment by Boundary Test

Abstract

The invention discloses a multi-agent system fault detection method based on the comprehensive judgment of the side-check. Each agent in the multi-agent system is taken as the apex, and the cooperative measurement relationship between the agents is used as the side to form a side-check network. Utilizing the functional relationship between edge attributes and Agent vertex attributes, it detects whether the attribute measurements of two vertices associated with an edge are normal based on the inspection of edge attribute measurements. Using the multilateral inspection results in the network, combined with the conditional probability of the multilateral inspection results calculated based on the false alarm rate and missed alarm rate of the unilateral inspection, the failure status of each agent is comprehensively judged according to the maximum likelihood principle. This method uses the cooperative measurement among multiple agents to confirm whether the attribute measurement of each agent is faulty, and can detect faults that are not easy to detect only by the agent's own measurement; its maximum likelihood judgment method is more accurate than simple voting; It can guarantee complete detection of any vertex fault whose number is less than the minimum number of neighbors of the vertex, so the choice of edge attribute measurement topology has a clear basis.

Description

technical field [0001] The invention relates to the technical field of multi-agent system fault detection, in particular to a method for detecting multi-agent system attribute measurement faults through comprehensive judgment of multilateral inspection. Background technique [0002] A multi-agent system is a unified abstract expression of various distributed systems that complete tasks through group collaboration. There are many different examples because of the different specific objects of Agents, such as unmanned aerial vehicle swarms, satellite formations, robot clusters, etc. Cluster systems, and distributed detection systems such as wireless sensor networks. With the rapid development of technologies such as artificial intelligence, sensors, network communications, and device miniaturization, it has become a trend to use a large number of micro-agents to work together, which can bring many advantages such as performance improvement, reliability increase, adaptability e...

AI Technical Summary

Problems solved by technology

The multi-Agent cooperative fault detection method essentially uses the measurements of different Agents and their relative or cooperative measurements to verify each other, so only the cooperative detection (that is, unilateral detection) between a pair of Agents can only find faults. But the source of the fault cannot be identified

In the use of multi-pair agent cooperative measurement synthesis (that is, the comprehensive judgment of multilateral detection results) to identify fault sources, the current research uses a simple minority-obedient majority voting method, without considering the error probability in unilateral detection. The quantitative impact of multilateral comprehensive judgment does not consider the impact of cooperative detection network topology on fault detectability, and it cannot guarantee complete detection of faults

Images