Fault localization in motorcycles using wavelet packet energy distribution

Motorcycles produce sound signals with varying temporal and spectral properties under different working conditions. These sound patterns can be used as an important source of information for automatic diagnosis of faults in motorcycles. Fault localization is a process of identifying the exact source of failure from a set of observed fault indications. The work proposed in this paper demonstrates the mechanism for identifying source of faults in motorcycles based on the distribution of energy in wavelet packet coefficients of the sound signal. The percentage of energy distribution among the nodes forms the feature vectors. The feature vectors thus obtained are used to train the neural network classifier. Performance of the proposed method is evaluated against six different types of faults. The experimental results show 95% accuracy for fault source localization in motorcycles.