The Search Of The Wavelet Neural Network Fault Diagnosis Method Based On Steel Cord Belt Magnetic Memory Signal

The steel cord belt is widely used in mining, metallurgy, construction, electricity and other departments, when the steel cord belt occurrence the fracture accident will bring significant economic losses and personal injury. This paper describes a non-destructive testing method of domestic and foreign steel cord belt failure. Study the failure mechanism of the steel cord belt and the metal magnetic memory testing principle. It shows that the wavelet combined neural network fault diagnosis method in fault diagnosis problem is advanced and the feasible and advanced in the steel core belt fault diagnosis.On the basis of the above, proposed the innovation method about combination of magnetic memory techniques and wavelet neural network fault diagnostic techniques for the steel cord belt fault diagnosis, provides a new feasible method for the steel core belt fault early diagnosis.This paper based on the metal magnetic memory testing signal fo the steel cord conveyor belt. Use the wavelet analysis technique for noise reduction preprocessing for the steel core belt of metal magnetic memory signal, extract the denoised magnetic memory signal feature vector as the input vector of a neural network fault diagnosis system, and then through the sample vector establish the BP neural network fault diagnosis system to identification and diagnosis the steel cord belt magnetic memory signal state. Finally simulation results show that the fault diagnosis system has good fault diagnosis success rate. Wavelet and neural network fault diagnosis method based on the detection signal of the magnetic memory to achieve the early diagnosis of the steel cord belt, can prevent the major accidents to occurrence.