Skip to Main Content
In order to diagnose the faults of the valve and the piston-connecting rod of internal-combustion engine (ICE), the vibration signals under normal and abnormal models were measured by experiments. Through continuous wavelet transform (CWT), the scale-wavelet power spectrum (SWPS) of signals was obtained. The wavelet power (WP) distribution on different scales of each model is observed to be similar and mainly concentrated in particular scope of 1~32. By analyzing the diversity of SWPS distribution, the WP that is most sensitive to the characteristic of each model were extracted by rough set (RS) theory as feature and taken as input to train the back-propagation neural network (BPNN). By the trained BPNN to diagnose the fault signals under detection, the correctness rate is 100%. The fault diagnosis method based on the integration of the SPWS, RS and neural network demonstrates to be efficient and feasible. It has preferable engineering applicability and referenced value to diagnosis for complex machines.