FAULT DIAGNOSIS OF ENGINE BASED ON CHAOTIC FEATURES AND SUPPORT VECTOR MACHINE

In order to resolve the complicated problems of the engine fault diagnosis,the vibration signals of the engine were obtained by simulating the engine valve faults,and meanwhile 3 chaotic features such as the correlation dimensions,maximum Lyapunov exponent,Kolmogorov entropy and their statistical features were obtained as the fault features.Different fault patterns were identified using support vector machine.The results indicate that the classification performance of the single chaotic feature is more poor,and with statistical features in tandem with chaotic features as the feature vector of fault signal,the performance of fault diagnosis is better.