A Two-Stage Fault Diagnosis Method With Rough and Fine Classifiers for Phased Array Radar Transceivers

Abstract

Transceivers are critical components of phased array radar (PAR) systems, and accurate fault diagnosis is essential for ensuring their reliability. However, many transceiver faults exhibit similar characteristics, making them difficult to identify. To address this challenge, a two-stage fault diagnosis method employing both rough and fine classifiers is proposed for PAR transceivers. In the first stage, a weighted support vector machine serves as the rough classifier to effectively separate easily distinguishable faults. For more complex faults that remain ambiguous, Fisher’s discriminating ratio is used to identify the most significant monitoring variables, refining the analysis further. In the second stage, a sparse momentum deep belief network (DBN) is developed as the fine classifier to accurately identify these challenging faults. The configuration parameters for both classifiers are optimized using a modified equilibrium optimizer to maximize performance. The proposed method is validated using a real-world dataset of PAR transceivers, with test results demonstrating superior accuracy compared to several existing intelligent diagnostic methods.