Demodulation is one of the most useful techniques for the fault diagnosis of rotating machinery. The commonly used demodulation methods try to select one sensitive sub-band signal that contains the most fault-related components for further analysis. However, a large number of the fault-related components that exist in other sub-bands are ignored in the commonly used envelope demodulation methods. Based on a weighted-empirical mode decomposition (EMD) de-noising technique and time–frequency (TF) impulse envelope analysis, a multi-scale demodulation method is proposed for fault diagnosis. In the proposed method, EMD is first employed to divide the signal into some IMFs (intrinsic mode functions). Then, a new weighted-EMD de-noising technique is presented, and different weights are assigned to IMFs for construction according to their fault-related degrees; thus, the fault-unrelated components are suppressed to improve the signal-to-noise ratio (SNR). After that, continuous wavelet transformation (CWT) is adopted to obtain the time–frequency representation (TFR) of the de-noised signal. Subsequently, the fault-related components in the entire frequency range scale are calculated together, referring to the TF impulse envelope signal. Finally, a fault diagnosis result can be obtained after the fast Fourier transformation of the TF impulse envelope signal. The proposed method and three commonly used methods are applied to the fault diagnosis of a planetary gearbox with a sun gear spalling fault and a fixed shaft gearbox with a crack fault. The results show that the proposed method can effectively detect gear faults and yields better performance than other methods.

High power density in-wheel motor could be achieved by integrating a high speed ratio (larger than 25) compound power-split mechanism (CPSM) with small motors. However, due to the exhaustive searching method adopted by the traditional lever analogy method, it is time-consuming to design high speed ratio compound power-split mechanism configurations satisfying high power density of in-wheel motor. In this paper an improved lever analogy method is proposed to find the optimal configurations with high speed ratio to satisfy the high power density in-wheel motor. In this method, a judgment algorithm about the rank of structure matrix is proposed to identify 3-node compound lever models of the CPSM. The improved lever analogy method can filter out useless configurations which significantly improve the calculation efficiency. The simulation results show that the calculation efficiency is improved by 215 times higher than that of the lever analogy method. Finally, 16 reasonable and 14 new configurations are obtained. This indicates that the improved lever analogy method can provide an effective way to design the high speed ratio CPSM which is widely used in-wheel motor driven vehicles.

No abstract

No abstract

No abstract