Application Of Data Mining In OPGW Lectotype

Data mining is a rising multidisciplinary technology and has begun to be focused, which is developed accompanying by the great deal of increased data and market competition urgent needs of information and knowledge. Data mining can automatically and expediently access the methods of knowledge, often called knowledge discovery from database.These patterns are hidded or recorded in large database, data warehouse, web and other huge information base or data flow. At present, information and knowledge acquired can be widely used in many fields, such as business management, production control, market analysis, engineer design and science exploring, and so on. The common categorizing technologies of DM include neural network, categorizing, clustering, text relevance analysing, mechanism analysis etc. In this paper, we try to lead the data mining technology to the type selection of the optical fiber composite overhead ground wire.In the high-voltage transmission line, using the OPGW as a communication medium of the distribution data could also be a cost-effective, advantageous, and effective way. Optical fiber composite overhead ground wire embedding the optical fiber unit in ground wire and composing the communication channel of electrical power system and combining the functions of grounding and data communications is a specialized power cable. The type selection of OPGW ensures the safety and the stability of power system including the match of electrical performance and the match of mechanical capacity.In this paper, the mechanism analysis is used to analyse the electrical performance of OPGW, giving the biggest short circuit fault of the OPGW type selection. By means of analysis the related factors of OPGW current and formula derivation, one kind of the match of electrical performance method based on impedance matching is presented. Through examples of analysis, the method is effectively used in the match of electrical performance of OPGW and provides the research object to the match of mechanical capacity.On the premise of electrical properties matching, the match degree between OPGW and ground wire determines whether OPGW can play ground and communications functions or not. The match of mechanical capacity is mainly embodied in stress and sag. Aiming at the calculation process of stress and sag is too complex, the model of stress and sag calculation based on BP neural network is proposed on the basis of synthetically analyzing the influence factors of stress and sag, combining with the march rule between transmission line and ground wire, OPGW and ground wire in this paper. Sub-network model of transmission line, ground wire and OPGW as well as corresponding typical training samples and testing samples were established respectively. The computation of stress and sag was carried out by using Matlab toolbox. Compared with the traditional method, the model given in this paper simplifies greatly the complexity of computation, which can accurate calculation the stress and sag up to proper selection of OPGW, it has certain engineering application prospect.Finally, according to the structural characteristics of OPGW and combining the structural parameters of ground wire, a series of candidate OPGW are designed. Again according to the match of electrical performance and the match of mechanical capacity, the best matching OPGW with ground wore is choosed and auto type selection comes true.