Application of Image Recognition Technology based on Fractal Dimension for Diesel Engine Fault Diagnosis

Abstract

A new method of diesel engine fault diagnosis that uses image recognition technology based on fractal dimension is proposed. The Wigner-Ville distributions of six kinds of vibration acceleration signals which are acquired from diesel engine cylinder head are calculated by time-frequency analysis, and a series of time-frequency gray images can be obtained from above distributions by image processing. According to the theory of fractal, we can obtain a group of fractal texture characteristic parameters from these gray images. At the same time, in the process of pattern recognition, we adopt BP neural network to classify these image texture fractal characteristic parameters, and then we can identify diesel engine valve gap abnormal status. Experiment results show that the proposed method can distinguish different texture characteristic of time-frequency gray images which are generated from different valve gap status of diesel engine, and this method is worth for further study.