Study On Circuit Fault Diagnosis Based On Neural Network Of Controller Of Main Engine Remote Control System

Main Engine Remote Control System(MERCS) is a system that controls the main engine through automatic control equipment in the Ship Central Control Room or Bridge. MERCS is essential to ensure the safety of sailing .It's fault can make the abnormal running of the main engine equipment,therefor leading to a serial of negative problems.As the core of MERCS,the controller is also very important.So it is particularly important to detect and remove it's circuit fault.Some traditional fault diagnosis technology can quite well solve the problem of fault diagnosis of some simple circuit,but do not work to large-scale and complicated electronic equipments such as controller of RCS.In recent decades , the rapid development of Artificial Intelligence(AI) , especially the application of Expert System(ES) ,Artificial Neural Network(ANN) and Fuzzy Theory in the area of fault diagnosis,opens up a new channel for intelligent fault diagnosis of electronic equipment. The fault diagnosis method based on ANN can realize the fault diagnosis to some complicated systems,which has the function of self-studying,and if combined with ES,is able to optimize the diagnosis knowledge of knowledge base and self-adaptively corrects the knowledge base according to the effectiveness of diagnosis results,thus improves the accuracy of diagnosis results and diagnosing efficiency to a rather big extent.In the background of project'Research on the comprehensive fault detecting device of Balaji and Losa system',after completely analyzing the circuit principle of the controller of MERCS,this thesis proposes an intelligent fault diagnosis method based on BP Neural Network and Expert System for large-scale circuit. This thesis firstly disgusses the principles of intelligent fault diagnosis,ANN,Expert System and signal processing technology,then according those principles designs the hardware and software of the fault diagnosing system for controller of MERCS,which adopts the fault diagnosing strategy of ANNES, and specifically discusses the circuit fault diagnosing mode built by BP Network.As results show in practical detecting process,compared with traditional circuit fault diagnosis method,the method this thesis proposes is very effectional and has the advantages of good self-adaptive characteristic,high accuracy and high efficiency,thus opening up a new way for traditional circuit fault diagnosis,bringing circuit fault diagnosis to intellectualization.To summarize,the subject this thesis studies has significant research value and the diagnosing system this thesis designs has great guiding meaning to practical application.