Robust Auto-Associative Kernel Regression (AAKR) Fault Prediction Method for Motor-Driven High-Voltage Circuit Breakers

Fault prediction of high-voltage circuit breakers is critical in enhancing system reliability and maintenance levels and therefore is a research hotspot. Most existing fault predictions for high-voltage circuit breakers suffer from single monitoring parameters, which do not conform to the complexity of actual equipment. Hence, this paper presents a comprehensive multi-parameter prediction approach for high-voltage circuit breakers and employs the evaluation results to optimize the operation of a smart grid. Specifically, this work proposes a fault prediction method based on robust auto-associative kernel regression (AAKR), which selects multiple parameters to construct a fault prediction model for high-voltage circuit breakers and using health data for fault prediction. The effectiveness of the proposed method is verified through a small amount of real fault data.