Design And Implementation Of Fault Diagnosis System Of Intelligent Computing

The development of modern industry,China's industrial equipment is also gradually refined,intelligent development,equipment integration is also higher and higher.If one of the parts fails,it may have a certain effect on the whole production process.But if one can grasp the working state of the system and its status is accurate identification,fault of related links for timely replacement of equipment,so that it can effectively reduce the fault caused by the loss.At present,the theoretical application of fault diagnosis is relatively rare in practical application.Along with the rapid development of intelligent computing,are applied to all fields of society the advantages of its application in the actual system is increasingly obvious,not only makes system more intelligent,also accelerate the computational speed.The fault diagnosis technology of intelligent computing is an important research direction in the field of pattern recognition in computer technology.Based on the fault diagnosis technology of intelligent calculation,the intelligent fault diagnosis system is designed and realized.The system provides a large number of bearing data as a data set for training fault diagnosis model and implement the user login registration,system management,data management,fault diagnosis model building and fault diagnosis function module.The subfunction that data preprocessing function of the data management module and fault diagnosis model building module are the core functions of this system.In data management module data preprocessing function,the system first uses the wavelet transform(WT)technology to reduce the noise of the original vibration signal;after noise reduction,the characteristics of the data are calculated;according to the selection principle of fault characteristics,the principal component analysis(PCA)is used for the reduction of time domain characteristics and the feature selection is completed.The pretreated data is used as the training sample set for the fault diagnosis model.In the fault diagnosis model building module,BP neural network algorithm is used and combined with Kalman filtering thought that the difference between theactual predicted value of the network and the theoretical expectation is fed back to the input layer,so this paper presents an information feedback BP neural network(IFBP)algorithm training fault diagnosis model.The information feedback BP neural network model improves the accuracy of the system diagnosis and solves the problem that the BP neural network is prone to local optimization.Through the system test,the system can accurately diagnose the bearing data,and can achieve the system's demand and expected design effect on both function and performance.