The Study Of Fault Intelligent Diagnosis For Rolling Mill Hydraulic AGC System

Hydraulic Automatic Gauge Control (HAGC) system is an important large-scale core system in strip mill. It is a key to determine the steel band quality and the yield. The HAGC system is a complicated, high-accuracy and synthetical control system with the characteristics of the mechanical system, electrical system and hydraulic system. As its structure and work principle are complicated, its faults have the characteristics of structural, parametric, non-linear, non-confirmation and so on, so the conventional fault diagnosis methods can't diagnose these faults quickly and correctly. It is urgent and necessary to research a practicle and feasible method to diagnose and exclude the faults of HAGC system intelligently, quickly and correctly.This paper researches the fault diagnostic strategies of HAGC system on the basis of generalizing abundant the domestic and abroad literature, combining with the modern fault diagnosis theory, artificial intelligent diagnosis theory, information fusion technology, database technology and computer technology.1. The fault types, mechanics and fail types of HAGC system are analyzed and researched profoundly. A new fault diagnosis method with facing the general performance analysis of system, combining with on-line monitor and off-line testing are presented. The fault diagnosis method based on multisensor information fusion is presented and applicated in HAGC system. The diagnostic frame with integrated neural network information fusion of HAGC system is builded. The fuzzy diagnosis of signal characteristic curve is applicated in HAGC system, the hard questions of decoupling faults and extracting the diagnostic parameters are solved, and the diagnostic accuracy is improved greatly.2. The basic principle and method with the multisensor information fusion technology are applied in hydraulic system on the basis of analyzing the characteristics and merits of information fusion technology. The method of integrated neural network information fusion is proposed, the fault diagnostic information will be used to diagnose faults effectively. The diagnostic frame with integrated neural network information fusion of HAGC system is builded, it will improve the confirmed diagnosis ratio of faults.3. The state characteristic signals of HAGC system are analyzed on the basis of expounding the signal analyzed methods of time-domain, frequency-domain and wavelet. Take example for electrohydraulic servo valve, combining with the fuzzy characteristic of HAGC faults, the fault diagnostic method based on fuzzy diagnosis of signal characteristic curve is applied in HAGC system, the fault states are identified correctly, so the method of feasibility and validity are approved.4. The on-line monitor and fault intelligent diagnosis software of HAGC system is developed with Visual C++6.0 lanuage and Access database technology. In this diagnosis...