Design And Optimization Of Reliability Evaluation Index System Of Electric Power Communication Network

With the progress of the technology and the development of the society, it is imperative nowadays to improve the degree of automation to meet the demands of smartgrid, and therefor electric power communication is getting more and more important in the power system. The reliability of the electric power communication network (EPCN), which supports many important services of electric power system, will have great effects on the safe, steady, economic and high-quality operation of the smartgrid.To realize the reliability assesment of an EPCN, a complete specification should be formulated to act upon. In this paper, based on the analysis of communication demands of electric power system communication services on the EPCN, a reliability evaluation index system, which can provide a basis for the work of reliability evaluation, is designed for the EPCN.This thesis firstly analyzes and summarizes the specific requirements on the EPCN from various aspects of demands of electric power system communication services, based on the related national standards, industrial standards in electric power and communications, the actual development situation of the EPCN and the maintainance experience. Based on the analysis and summarization, formulate the design scheme of index system and establish the preliminary reliability evaluation index system of the EPCN. Secondly, through the analysis of the common existing problems of index system, work out the relevant index system optimization scheme, and optimize the reliability evaluation index system of the EPCN. The main tasks of this thesis include: Checkout and optimize the availability and reliability of the index system on the basis of relevant theories and methods in mathematical statistics; Calculate the weight of each index with AHP (Analytic Hierarchy Process) and consistentmatrix methods and optimize the index weight by correlative coefficient approach to eliminate the negative effect of index correlation; Obtain the index contribution value by neural network analysis, take the index weight value into consideration and get rid of the redundant index to reduce the index system redundancy. Finally, after the design and optimization, an evaluation system platform is developed using ASP.net (C#) technology and MSSQL for database.