The Research of Fault Diagnosis for Gasoline Engine Based on WP-GA -NN

In this paper, a fault diagnosis method based on wavelet analysis and neural network combining loosely has been brought. With the wavelet packet decomposition results of the signal as a neural network input value, using the genetic algorithm to optimize the parameters of neural network globally, and finally we can use the trained neural network for fault diagnosis. The simulation results show that this method has a higher computing speed and accuracy than the quasi-Newton algorithm. The method is applied to automobile engine fault diagnosis, and the result confirmed its feasibility and effectiveness.