Noise Source Identification of the Electric Bus Powertrain Using a Wavelet Transform and EEMD-RobustICA

Electric buses have been used widely as cities' short-range commuter vehicles, because of their excellent power, fuel economy and emission characteristics. However, the lack of a noise masking effect for the traditional internal combustion engine, the high-frequency noise becomes more prominent for the powertrain system. The high-frequency noise gives people an unpleasant feeling on psychological and physiological. To control electric vehicle powertrain noise, the identification of the main noise source of the powertrain is well needed. In this paper, Empirical Mode decomposition (EMD) combined with Independent component Analysis (ICA) and continuous Wavelet transform (CWT) was used to identify the main noise source of the electric bus powertrain. The contribution of each noise source to the overall noise level was calculated and compared. The results showed that there were four main noise sources of the electric bus powertrain: the engagement noise of the shifting gear and constant-mesh gear, cogging harmonic noise and switching frequency noise. The contribution of the shifting gear engagement noise was the greatest, followed by the constant-mesh gear engagement noise, and the contributions of the cogging harmonic noise and switching frequency noise were approximately of the same degree.