Fault Diagnosis Method for Automotive Panel Processing Equipment Hydraulic System Based on Rough Sets - Neural Networks

Abstract

The application, which the intelligence is used in the fault diagnosis, is the main direction of research currently, especially in the fault diagnosis of large mechanical equipment. In order to improve hydraulic system failure diagnosis of high-speed deep drawing hydraulic press, reduce the efficiency and accuracy of difficulty diagnostic staff. By using rough sets theory combining neural network and the method of large NC hydraulic press, hydraulic system fault diagnosis of diagnosis. This paper established based on rough set - neural network fault diagnosis model, and the following hydraulic cushion hydraulic system as an example, the diagnosis in establishing the fault table based on the rough set theory to fault table attribute reduction and generating rules, will rule input to the BP neural network was trained learning. Get in neural network after the test data repository and simulation. Test results show that the method for the diagnosis of high-speed deep drawing hydraulic press hydraulic system fault is effective.