Fault Localization of Industrial Robot System based on Knowledge Graph and Bayesian Network

Abstract

In order to solve the problems of diversified fault data, low efficiency of diagnosis methods, and low utilization of fault knowledge in industrial robot systems, this paper puts forward a fault localization method for industrial robot systems based on knowledge graph and Bayesian network. Firstly, the fault knowledge graph of industrial robot system is constructed based on text data, and the industrial robot fault knowledge graph is stored in Neo4j database; Then, the knowledge graph is mapped to a Bayesian network, and the transcendental probability of root node is obtained by using triangular fuzzy function as the transcendental probability of leaf node occurrence probability; Finally, the probability importance and key importance are added, and the three indicators are comprehensively compared to locate specific fault events efficiently and accurately. Experimental results prove the validity and feasibility of the method, which can improve the accuracy of fault location.