Research On Fault Diagnosis System Of The Elevator Based On Fuzzy Neural Network

The market demand of elevators keeps increase as the rapidly development of modern cities,elevators have brought unprecedented convenient for people's daily life. People require the elevator control system to be of higher safety and reliability,and it is especially important to find and eliminate elevator controller's fault.As the computerized and high-performance of the elevator control system,the requirements of the professional technique and testing methods mean a corresponding increase in the inevitable.So we can only equip with advanced testing equipment in order to guarantee better fault diagnosis on elevators.According to the current development of Elevator control technology,based on CAN bus for the elevator controller,the paper would do research on fault diagnosis of the elevator controller which apply a combination of fuzzy and neural network theory, and design a intelligent and portable elevator fault diagnosis cards which can be realized real-time diagnosis of elevators' fault.Main research content are as follows:Firstly,theoretical analysis for the consistent of the elevator controller and its work principle.Received the characteristic of elevator controller's fault by FMEA(Fault Model and Efficient Analysis).the structure of the diagnosis system was set up,and the fault tree was built.Explained with examples the fault of elevator controller was analyzed by the analytic method of fault tree.Sencondly,expound the basic theory on fuzzy and neural network respectively, and introduce this two theoretical techniques used in fault diagnosis.Then show the essentiality for a combination of both and the mode of combination.Based on the intelligent control system of the elevator which has already been exploited,it has accomplished the design of hardware circuit for the fault diagnosis system of elevator.Including the design of signal acquisition,liquid crystal display circuit and communication interface.Finally,a fault diagnosis model that combine the fuzzy theory and neural network was chosen to apply to elevator fault diagnosis.Set up the fault diagnosis expert system for elevator.It is theoretical analyze the method of knowledge acquirement,the method of knowledge denotation,the method of knowledge inference and the method of knowledge explain,and database design,management and maintenance.Focuses on using the production rules and fuzzy and neural network method to research on fault diagnosis of the elevator.