Automobile Fault Diagnosis System based on Improved Neural Network

With the widespread use of automobiles, automobile engine fault diagnosis system is of great significance. According to the analysis of engine fault type and fault symptoms, 11 kinds of engine typical fault symptoms and fault reasons are extracted. Fault information is taken as training sample of neural network after the normalization. The optimized BP neural network based on artificial fish swarm algorithm is used in automobile engine fault diagnosis. The experiment results show that the improved BP neural network has good generalization ability and can effectively deal with different types of faults. The results are accurate and reliable, which can provide important reference for automobile fault diagnosis system design.