Condition Monitoring of Railway Vehicle Suspension System Based on PCA-SVM Method
摘要
The suspension system is critical to ensure the running safety and comfortability of railway vehicle. This paper employed conventional machine learning method of principal component analysis and support vector machine (PCA-SVM) to diagnose the damper fault, wheel surface fault, roller fault, damper fault coupled with wheel surface fault and damper fault coupled with wheel and roller surface faults. The effectiveness of this method was verified by data obtained from a 1/5th scaled roller rig. The results shown that the performance of PCA-SVM was acceptable for railway vehicle suspension system monitoring.