Research On Remote Intelligence Fault Diagnosis Technology For Complicated System

The Complicated System plays an essential role in the national economy and the mass daily life. As a kind of typical complicated system, the Railway Traffic Equipments (RTE) are exerting greater and greater influence on people, and thus the research on Remote Intelligence Fault Diagnosis of RTE is significant to improving its safety, reliability and validity, to reducing the economic expense and to ensuring people's security. This paper takes RTE as the research object, from the angle of general technology, support technology, core technology and realization technology, making a research on the Remote Intelligence Fault Diagnosis. The main contents of this paper are as following:Firstly, the structure of the Remote Intelligence Fault Diagnosis is established. To effectively realize the information collection, integration, transmission, analysis, judgment and decision, this paper proposes three-leveled diagnoses, which are on-the-spot diagnosis system, local diagnosis center diagnosis system and remote center diagnosis system. Their working principles, respective logic structures and function models are all presented in details. This paper also analyses the effective integration of diagnosis system and corporate information system.Secondly, the LON(Local Operation Networks) fieldbus, the wireless access and the multimedia management of information system are regarded as the support technology, and their application in the Remote Intelligence Diagnosis System are studied in this paper. In analysis of the characteristics of the above three sorts of technology and combining the requirements of RTE, the LON-RS232 gateway has been developed. This paper presents a design project of remote diagnosis system with the adoption wireless access technology, studies some key technology of the multimedia information management system, and analyses its function model.Thirdly, the conception of hybrid structure tree (HST) is presented. HST is a kind of platform for constructing the intelligence diagnosis system. This paper brings out the basic model of HST, introducing its diagnosis principle with a detailed example and gives the knowledge encoding algorithm which can organize and manage knowledge effectively.Then, the intelligence diagnosis system based on HST is studied, and the emphasis is put on the following three key questions: (1) To put forward a new approach of knowledge acquisition, which integrates the rough sets and experts' experience, and introduce properties importance and rule certainty factor to rough sets; (2) To research the diagnosis reasoning based on fuzzy sets, and present two fuzzy reasoning methods; (3) To design the intelligence interface which possesses the ability of natural language understanding, and direct the algorithm of word segmentation as well as the computation of sentencessimilarity.Fourthly, based on the fruits of theoretical research, the Intelligent Control and Remote Monitoring Diagnosis System (ICRMDS) is developed, which can provide RTE with a kind of auxiliary, original, distributive and intelligent remote monitoring, diagnosis and maintenance system. The successful test of ICRMDS proves the correctness and feasibility of the theories of this paper. Furthermore, ICRMDS has been applied to practical use.Lastly, all the researches of the paper are summarized and what should do next is listed alike.