A comprehensive electromagnetic compatibility analysis method for electronic systems based on knowledge graphs

Abstract

The present invention relates to a method for electromagnetic compatibility analysis of an integrated electronic system based on a knowledge graph, belonging to the field of electromagnetic compatibility. Electromagnetic sample data is obtained through actual measurement and engineer experience documents, etc. A machine learning-based method is used to perform entity recognition and entity and attribute extraction on the sample data to obtain the entities, relationships, and attributes required for the electromagnetic compatibility knowledge graph of the integrated electronic system, and a knowledge graph of the electromagnetic compatibility of the integrated electronic system is constructed. Then, the information of the knowledge graph is imported into a hierarchical Bayesian network, and probabilities are calculated through the weight matrices of each component in the internal module, and a knowledge graph for predicting interference and failure probabilities is generated, so as to realize electromagnetic fault diagnosis and interference prediction of the integrated electronic system. The present invention aims at the electromagnetic compatibility problems inside the chassis of the integrated electronic system, intelligently extracts relevant data on the electromagnetic compatibility of the device, and can solve the problems of electromagnetic interference fault diagnosis and prediction.

Images