Research And Application On Fault Diagnosis For Loading Machine Based On Neural Network

This paper research on loading machine fault diagnosis based on neural network. Its purpose is to promote the high-tech content and market competition ability of loading machine through loading machine fault diagnosis using loading machine signal process,fault feature extraction and neural network.This paper begins with the analyzing of existing methods for fault diagnosis and their advantages and disadvantages. Then it designs the process of loading machine fault diagnosis and gives the functions and detail designs of some important components in the process. It extracts fault features based on analysis of the loading machine signal. After that it founds two neural network models. One is BP neural network model; the other is assembled neural network model. Compared with the performance of the two models, this paper selects the fit one for loading machine fault diagnosis. Finally it designs and implements a software system named loading machine fault diagnosis system, its function includes data collection,signal preprocess,fault extraction,fault diagnosis,neural network training,knowledge management and user management. At the last part of this paper is the conclusion followed by future works.