Aircraft Engine Multi-Condition Detection Method Based on Single Classification Limit Learning Machine Algorithm

This paper provides a method to detect multiple working conditions of aircraft engines based on single category limit learning machine algorithm. The k-means clustering algorithm is used to realize automatic division of working conditions. By building detection models under different working conditions, the parallel monitoring of multiple models is realized. The semi supervised single category limit learning machine algorithm is used as the anomaly detection algorithm. By building the normal domain of complex data sets, The abnormal index of the equipment is calculated according to the output deviation of the sample to be tested. In addition, the method of moving average filtering and standardization is adopted to deal with noise and dimensional problems, and data preprocessing is completed. This paper is a complete multi working condition anomaly detection system for the aircraft engine system. The degradation status of the engine is obtained through the obtained anomaly indicators, and the anomaly warning is realized before the equipment failure, which ensures the safety and reliability of the aircraft operation.