Diesel Engine Gearbox Fault Diagnosis Based on Multi-features Extracted from Vibration Signals

Abstract

Abstract Engine vibration is the consequence of a combination of internal forces and external forces, and the vibration signals contain a variety of information, also it is easy measurement, low cost and strong robustness. Hence, one of the most common and promising methods in fault detection area is vibration analysis which is effectively used in diesel engine fault diagnosis. In this paper, a new fault diagnosis method for diesel engine with vibration signals analysis and feature extraction based on ensemble empirical mode decomposition combined with support vector machine was proposed. Firstly, the vibration signals are first preprocessed and the intrinsic mode functions containing fault information was obtained by ensemble empirical mode decomposition. Then, extracting a feature vector from the intrinsic mode functions, time-domain digital features combined with Hilbert marginal spectra as a feature vector of gearboxes. Next, the fault classification of gearboxes were identified by support vector machine. Finally, the effectiveness of the proposed method was proved through the comparative experiments based on a large number of samples from rotor laboratory bench, the results illustrate that the proposed method could identify the multiple faults information from the collected vibration signals under different system states of gearboxes accurately and effectively.