Neural Network Information Fusion Fault Diagnosis Methods Research And Application In NC Machine Tool

The information fusion technology can comprehensively handle the information frommulti-sensor，and it has been widely applied in many fields.In the fault diagnosis of NCmachine tool, a multiple sensor fault diagnosis integration system model is preseneted,and it is applied to the NC machine tool failure diagnose. Based on in-depth analysis ofthe structure and fault mechanism of the NC machine tool, the methods of fault diagnosis，the wavelet packet transform is used to remove the noise in signal and extract the faultfeature vector. In addition, it builds the feature vector of signal from NC machine tool.That all laying a foundation for further studying the method of neural network informationfusion fault diagnosis.The structure of neural network and learning algorithm，and it is applied to the faultdiagnosis, and puts forward a new fusion fault diagnosis model is introduced. It is thefeature-level fusion diagnosis. On the basis of fuzzy technology combined with neuralnetwork. It puts forward an improved structure of fuzzy neural network and builds thecorresponding model framework of fault diagnosis fusion.The wavelet neural network isintroduced wavelet into neural network. Thereby the wavelet neural network not only keepthe mufti-resolution characteristics of wavelet, but also keep the self-learning and patternrecognition capabilities of neural network. It replaced the activation function of artificialneuron with wavelet function to lead the characteristics of mufti-scale into neural network.The experiment results show that it can greatly improve the diagnostic accuracycompared to the general neural network fault diagnosis methods and verify the feasibilityand effectiveness of this method.