Application of a Bayesian Network Based on Multi-Source Information Fusion in the Fault Diagnosis of a Radar Receiver.

A radar is an important part of an air defense and combat system. It is of great significance to military defense to improve the effectiveness of radar state monitoring and the accuracy of fault diagnosis during operation. However, the complexity of radar equipment's structure and the uncertainty of the operating environment greatly increase the difficulty of fault diagnosis in real life situations. Therefore, a Bayesian network diagnosis method based on multi-source information fusion technology is proposed to solve the fault diagnosis problems caused by uncertain factors such as the high integration and complexity of the system during the process of fault diagnosis. Taking a fault of a radar receiver as an example, we study 2 typical fault phenomena and 21 fault points. After acquiring and processing multi-source information, establishing a Bayesian network model, determining conditional probability tables (CPTs), and finally outputting the diagnosis results. The results are convincing and consistent with reality, which verifies the effectiveness of this method for fault diagnosis in radar receivers. It realizes device-level fault diagnosis, which shortens the maintenance time for radars and improves the reliability and maintainability of radars. Our results have significance as a guide for judging the fault location of radars and predicting the vulnerable components of radars.