Research On Intelligent Fault Diagnosis Technique Of Complex Equipment

Fast and accurate faults detection and diagnosis is an important part of battle effectiveness of equipment. A certain type of equipment is an integrated and complex system with optical, mechanical, electrical components, and most of its faults are relatively complex and uncertainty. The accessorial equipment for faults detection and diagnosis only has the function of performance test and simple fault judgment, can not supplies comprehensive and accurate fault status for the equipment. In addition, the accessorial equipment lacks of autonomous learning and updating function, for its knowledge is relatively fixed and dated.Aiming at the actual demand of complex equipment fault diagnosis, this paper made the fault intelligent diagnosis technology of a certain type of equipment as research object, analyzed the characteristics of equipment and its faults, presented a intelligent diagnosis method which was based on BP neural network and fuzzy theory, studied the key techniques for realizing intelligent diagnosis method systematically and deeply, such as diagnosis model design, software realization, fault knowledge database, self-study and human-computer interaction and so on. The main research contents are list as follows:(1) Research on fault diagnosis object and establishing fault database. This paper formulated fault knowledge data formats and establishing the fault database based on the analysis of equipment composition and working principle; this paper also analyzed the rules and characteristics of fault samples, especially the nonlinear and inaccuracy mapping relationship fault symptoms and fault causes; in addition, this paper presented fault diagnosis project of the complex equipment.(2) Method and technique research on fault intelligent diagnosis based on BP neural network. Aiming at complicated nonlinear mapping relationship of fault symptoms-causes and self-study capability, this paper introduced the BP neural network method into fault diagnosis, constructed intelligent diagnosis model based on BP neural network, studied technique approach for model realization.(3) Method and technique research on fault intelligent diagnosis based on. Aiming at inaccuracy mapping relationship of fault symptoms-causes, this paper introduced the fuzzy comprehensive evaluation method into fault diagnosis, constructed intelligent diagnosis model based on fuzzy theory, studied technique approach for model realization.(4) Software realization for fault intelligent diagnosis model of a certain type of complex equipment. Based on the research of fault knowledge and intelligent diagnosis technique, this paper adopted object-oriented programming ideas and Microsoft Access 2003 as database, programmed software system for fault intelligent diagnosis of the complex equipment in Microsoft VC++ 6.0 programming circumstances.(5) Experiment and analysis of software system function and performance. This paper tested the software system for fault intelligent diagnosis of the complex equipment with fault samples, analyzed diagnostic function and performance of the software system.