Font Size: a A A

Design And Implementation Of The Diagnosis System Based On The Alarm Data Of Electric Power Operation And Maintenance

Posted on:2017-04-14Degree:MasterType:Thesis
Country:ChinaCandidate:X F LiuFull Text:PDF
GTID:2272330488484406Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet technology, the network has become an important and indispensable part of all walks of life. At the same time, the constraints due to the increasingly large scale of managed network, its requirements for heterogeneous and dynamic, the complexity of the management functions, user demand for high-quality multimedia services, the lack of experienced management person and many other factors, it makes traditional network management, in terms of performance or function, are unable to meet the management of the various existing large and complex network systems. Therefore, it is necessary to introduce artificial intelligence technology into network management, which is intelligent network management.In this paper, the design and implementation of a diagnostic system based on the power of the operation and maintenance of alarm data, which can simulate human experts to solve a class of problems, which applies artificial intelligence and computer technology. It base on the knowledge of experience of one or more experts in their field, to make reasoning and judgment, finally achieve purposes of independent decision-making and provide solutions. In this paper, the knowledge base is built by data mining technology, the algorithm of data mining, Apriori. It is improved from three aspects for achieving dynamic mining, improving knowledge generation rate and generating knowledge automatically. First, increase the weight to make the rules meaningful. Then, filter data before connecting to reduce the number of candidates, improve the efficiency of the algorithm. Finally, we can take advantage of existing result of old mining to update the knowledge base when new affairs come, rather than run mining algorithms repeatedly. Using inference engine technology and new storage, we can improve and simplify the Rete algorithm for eliminate memory affect due to different conditional sort order, and increase the efficiency of delete rules. Eventually, using object-oriented approach and the Java programming language, combined with the thought of MVC framework and the background of data of operation and maintenance of the power system, this paper implements a fault diagnosis system, to achieve the pre-sentence, reasoning and relevant contact display of failure.
Keywords/Search Tags:diagnosis systems, data mining, Apriori algorithm, Rete algorithm
PDF Full Text Request
Related items