Earticle

The life prediction method based on the power spectral envelope energy and Support Vector Machine (SVM), which is the combination of time-domain and frequency-domain, is proposed for the key component of aircraft’s (aircraft air refrigerator) life unpredictable problems. By first, made a long-term test for aircraft air refrigerator through aircraft-specific life testing platform, then got a lot of acceleration parameters which characterize the health status of the aircraft air refrigerator. Through the analysis of acceleration parameters’ power spectrum, its power spectrum envelope energy would be obtained. Used the power spectrum envelope energy to construct feature vector, and designed SVM life prediction model and using this model to predict aircraft air refrigerator’s lifetime. The research results show that the life prediction method based on the power spectral envelope energy and SVM can achieve a good life prediction of aircraft air refrigerator, and laid the foundation for life prediction of air refrigerator in aircraft engineering application.