Research on Fault Diagnosis of Diesel Engine Based on Bispectrum Analysis and Genetic Neural Network

Abstract

Aimed at the problem of the feature of nonstationary vibration signal can not be extracted easily; the bispectrum analysis technique was researched in this paper. Bispectral characteristic frequency faces were searched along the parallel to the diagonal line at certain step in the bispectral modulus field, and the mean magnitude was calculated to get the feature parameters. The method was applied in fault feature extraction of piston-pin of diesel engine successfully. Finally, the piston-pin faults were diagnosed associated with genetic neural network.