Design And Research Of Machine Fault Diagnosis System

Machinery is widely used in the modern mass production. As the mechanical equipment is more complicated and large-scale due to the industrial production and the development of science and technology, traditional machinery diagnostics can not meet the need of actual diagnosis. Continuous development and improvement of the artificial intelligence technology helped to promote the reform of mechanical diagnostic technology, as a result, the study of combining artificial intelligence with mechanical fault diagnosis is of vital importance.Starting from the actual situation of production and directed against the complexity of equipment working environment under normal condition and the actual situation of insufficient diagnostic experts in most production unit, the writer analyzed the relative theory of combining the expert system and mechanical fault diagnosis. The main part came using the asynchronous motor which is widely used in production as the diagnostic object; the writer designed and studied the motor fault diagnosis expert system, and analyzed the key technologies of its structure and realization.At first, the thesis gives an introduction to the structure of the mechanical fault diagnosis expert system, using the asynchronous motor as the diagnostic object, analyzes and studies its structure and fault mechanism, and designs the diagnostic expert system with rules and frames. Acquiring knowledge through fault tree analysis, the establishment of knowledge base is the main point of the thesis. Mixing the knowledge of fault tree with the frame representation, and minimal cut sets and rules of integration, this kind of knowledge can not only describe the knowledge composition of the motor but make the design of the inference engine more concise.Another research emphasis of the fault diagnosis expert system is the implementation technique of the inference engine. The thesis designs an inference engine with forward reasoning as its inference direction and depth-first search as its search strategy. If the fault knowledge conflict occurred, we use a conflict solution in which we arrange the fault cause in the order of importance; this solution can help determine the fault position and the diagnosis result efficiently.Ado database access techniques is used, the database and knowledge base to create and manage, the fault tree is integrated into knowledge base management module, users can understand the knowledge structure and specific problems of knowledge base to update and management. At last, the Visual C++ 2005 program development tools is used; the Access 2003 database finished the design of the asynchronous motor failure diagnostic expert system, and its effectiveness and availability was proved by the test on the motor experimental platform.