Multilevel decomposition of vibration signals from motorcycles using the wavelet–Hilbert transformation

Abstract

To maximize comfort and safety it is important to control motorcycle vibration. In order to assess the vibration of motorcycles, measurements need to be taken from these machines. Vibration signals acquired using sensors mounted on the machine always include the excitation components from the engine, the road surface, and the structure of the motorcycle frame. Generally, these vibration signals are coupled together. Presently, there is no ideal method to separate these signal components to allow better analysis of the different contributing factors. Contributions to the total vibration signal from the main vibration sources (the engine and the road surface) cannot be determined exactly. To solve this problem, a multilevel decomposition algorithm based on the wavelet–Hilbert transform is proposed in this paper. The paper formulates the vibration signal model for a motorcycle and discusses the performance of the multilevel decomposition algorithm using simulation and experimental methods. The simulation and experimental results show that the proposed new algorithm is an effective method for the identification of vibration sources on a motorcycle.