The Research Of Remote Intelligent Monitoring, Diagnosis And Maintenance System For Complicated Equipment

With the development of technology, the structure of equipment is more complicated, itsfunctions are more powerful, some advanced technologies such as information and intelligentare used generally. It is difficult to diagnose the fault of equipment running everyday and keepits performance well.It is our purpose to provide a system that can realize remote monitoring diagnosis andmaintenance (RMDM) for complicated equipment effectively and intelligently. The maincontents of this paper are followed.Firstly, the architecture of RMDM system is established. It is composed of three levels: remotediagnosis center, local diagnosis center, maintenance on-the-spot. All of these resources includemanufacture factory, design and research organizations of equipment and personnel on-the-spotare integrated by Internet. So these resources can be shared easily. Then, we analyze the logicarchitecture of system in details, definite relations of each module, and build up function modelof this system.Next, the technology of modeling used to diagnose complicated equipment is researched. On thebasis of large system theory, we construct the graphical model of complicated equipment. Therelations and levels of each part in equipment are defined. And the process of fault propagatedwithin equipment can be decided. Lastly, the probable model is constructed.In the fourth chapter, we analyze Bayes Networks-ihe theory of uncertain reasoning. And it isimported into intelligent fault diagnosis and task decomposition process for complicatedequipment. The method have some advantage, such as reasoning in two-way (effect and result),diagnosis based on model, and using experts knowledge. ?In the fifth chapter, we set up knowledge model, node-calculation model on the basis of BayesNetworks theory, and demarcate task model, build up distributed cooperation diagnosis modelbased multi-Agent technology. The function module of agent is defined clearly, the informationpropagated in each module and relations among each module are analyzed in details. So we canuse advantages of all kinds of diagnosis resources, and make them cooperate to solve problems..The model is composed of components, and recombined easily as we need.Last, we develop the prototypical system, construct the environment of RMDM system running.And the most functions are realized in the prototype, it is very important beforehand research forrealizing this system under the environment of Internet finally.