Neural network parameter optimization and fault diagnosis based on orthogonal experiments

The wavelet analysis and neural network for fault diagnosis system in practice has become a hot research topic in the fields of pattern recognition system in recent years. Acoustic emission technology is used for vibrating screen´s fault diagnosis in this paper. The energy feature vectors of signals extracted by use of wavelet packet analysis is regarded as neural network input vectors, and the system parameters are optimized by means of orthogonal test method. The results show that the detection method can effectively determine the working state of the vibrating screen to achieve real-time monitoring.