Boring Machine Hydraulic System Fault Diagnosis Based On Fuzzy Neural Network Research

As a kind of special large machines used in tunnel,the roadheader is prone to failure under working adverse environment,and hydraulic system is one of the main failure parts of roadheader.The characteristics of plurality elements and complex loop and tightness lead to failure mechanism,influencing factors and performance in a variety of forms.Operator is difficult to quickly and accurately find out the cause of the problem and the correct fault processing,thus to bring personal security and economic security.Therefore,fault diagnosis of hydraulic system of roadheader is necessary. Because of the complexity of the roadheader hydraulic system, the traditional diagnosis method is difficult to meet the requirements, then the diagnosis must be adopted in the intelligent fault diagnosis methods.This paper studies a method of the combination of Fuzzy theory and BP neural network, which is used to fault diagnosis of roadheader hydraulic system. A fault diagnosis framework is constructed based on Fuzzy neural network and is realized in the form of software used VC++and Matlab. This paper mainly completed the following work:(1) The operation principles of the roadheader hydraulic systems is analyzed,and the commom faults of the hydraulic system and the character of the roadheader hydraulic systems are summaried by researching the failure mechanism and failure mode.By this,we select fault characteristic parameters that this paper needs,which prepare for this paper.(2) According to the expert system design principle,this paper establish the general structure of the knowledge base including fault type base, fault samples base,fault rule base and so on.The relational database is introduced to knowledge base system, using the technology,manages the knowledge base to lay the foundation of fault diagnosis expert system development.(3) BP neural network and fuzzy theory are introduced briefly and analyzed the advantages and disadvantages for fault diagnosis,and are combined. According to the actual situation of the hydraulic system,this paper established the BP neural network model and fuzzy BP neural network model respectively and uses MATLAB simulation.The results show that the convergence speed of fuzzy BP neural network is faster and its accuracy is higher,its performance is better than BP network,and suitable for roadheader hydraulic system fault diagnosis.(4) This paper puts forward the design scheme of the whole system software,and disigns each functional module structure and interface,develops roadheader hydraulic fault diagnosis expert system with VC++6.0and Microsoft Access database software platfor. Fuzzy theory determine the degree of fault,neural network complete self-learning function. The realization of the software verifies feasibility of knowledge base based on database technology and the fuzzy neural network reasoning mechanism...