The Research On The Intelligent Diagnosis Of Electromechanical System Based On Neural Network

Based on the analysis of the research in the faults diagnosis of electromechanical system and its application both in home and abroad at present, this thesis researches into the selection of observation modes, the collection of fault portents and the pattern recognition of system faults in the faults diagnosis of electromechanical system. It also introduces the basic principles of artificial neural networks and its applying methods during the collection of fault feature information and its mathematical modeling and algorithm in the faults diagnosis. An example of faults diagnosis in a complicated electromechanical system- horomill is cited; and according to its concrete features, this thesis covers the measuring and the collection of fault feature information, the composition of the software and the hardware of faults diagnosis system, and the collection of sample books. The author takes the algorithm of faults diagnosis of artificial neural networks to emulate the faults diagnosis of the horomill system. The thesis proves the feasibility and the superiority of faults diagnosis of the artificial neural networks-based electromechanical system.