Compactor Intelligent Fault Diagnosis System

This paper derives from the research and development of control system of intelligent road roller in "863" project. Sometimes, intelligent road roller integrates hydraulic technique, computer technique, microelectronic technique, and sensing and testing techniques, so the advanced technology and the complex structure add greatly to the difficulty of repair. Most of the time, airmada work together on the construction site, if the road roller occurs fault, the other construction machinery cooperate with it also must stop working, which will delay construction period and yield huge financial losses. The purpose of the research of this paper is just to achieve state monitoring, guarding against and diagnosing fault, and decreasing losses caused by road roller' s faults.Intelligent fault diagnosis system of vibratory road roller which consists of neural network and expert system is built in this paper by detailed analysis of structure and principle of vibratory road roller and comparison of main methods of building fault diagnosis system, this system can diagnose usual fault of vibratory road roller. In this paper, analysis and diagnosis of fault mechanism of YZC12Z intelligent road roller, overall design of the intelligent fault diagnosis system, realization of critical technology of the intelligent fault diagnosis system, and loading and debugging of the intelligent fault diagnosis system are discussed in this paper.There are some key technologies, such as knowledge library, ratiocinative mechanism, learning mechanism, expository mechanism, database and the inter-importing mechanism of neural network and expert system in this intelligent fault diagnosis system. The knowledge library, ratiocinative mechanism and learning mechanism all consist of neural network and expert system, and they should be treated separately. In knowledge library, neural network is used to indicate the surface knowledge, so as to instinctively associate with faults, while frameworks are used to indicate deep-seated knowledge to logically confirm the diagnosis. In ratiocinative mechanism, neural network adopt right directional reasoning method. And the ratiocinative mechanism of expert system contains two conditions, the validation to the diagnosing result of neural network and the alone ratiocination. Learning mechanism contains neural network' s study of expert experience and new faults.This system is achieved on the basis of controller and display produced by EPEC Company, software and hardware are closely linked, so the appropriate development language must be selected. It is attached by C on VC++ platform in this paper.A contrived fault is generated for the diagnosis system to analyze. And the final result approves that the intelligent fault diagnosis system achieves all the functions required.