Intelligent Fault Diagnosis For RH- KTB Vacuum System

RH-KTB vacuum system is complex-large system used to refine and purify steel.In practice the system often get out of order which effects product quality so far as to production efficiency.However most monitoring software of the system are not able to process state decision and fault diagnosis.When fault occurs,workmen have to manually find position of fault and removal of faults.Obviously it is inconvenience for production.Sometimes special position of fault and failure cause are uneasy to be found accurately.So it is important that an intelligent fault diagnosis system for RH-KTB is developed.In this paper species and features fault of RH-KTB vacuum system is studies and the fault tree is found against the system.The relationship amony faults is obtained,at the same time original information acquisition system is developed.The process conditions of RH-KTB demands for quick-diagnosis.In this paper,we have quick-fuzzy clustering analysis to this system.The original information system is reducted according to rough sets theory.The quick diagnosis means quick learning and classification.In this research decision tree theory is used to learn and classify reducts sets.Finally an intelligent fault diagnosis model based on rough sets-decision tree theory is put forward aiming at the high-duty vacuum metallurgical system of which the possible faults take place frequently and unwarnedly and need quick and exact fault diagnosis. The model is highly self-learning and self-organizing especially applicable for the fault diagnosis to heavy-duty and complicated vacuum systems.Theoretically,disturbed data can be eliminated according to the model. Describing the theory of rough sets-decision tree algorithm,the paper presents the procedure of rough sets-decision tree theory fault diagnosis,with an actual diagnosis process given as an example to show the intelligent fault diagnosis for RH-KTB vacuum metallurgical system.Thus,effectiveness of the algorithm is proved through an analysis of the exemplification.A RH-KTB intelligent fault diagnosis system based on the model is developed.