Forecasting Of Theory Line Loss Rate For Municipality A Network Based On Artificial Neural Network

The line loss rate of power system is an integrative economic and technical index. It not only reflects the rationality of network and running, but also reflects the technical and administrant level of power corporation. Of course, it also impulses correlative departments in corporation to adopt some measures to reduce line loss rate. Many scholars has done a lot of researches on line loss computing and gain prominent achievement at home and abroad, but the research on line loss forecasting method is few. In order to actively respond central government's policy of constructing abstemious society and make great efforts to reduce power transmission loss, the research on line loss forecasting became more and more important.On the basis of investigation on the related literature of computing method of line loss, a hybrid method used ANN's capability of learning and illation is proposed for the purpose of forecasting line loss rate in this dissertation.The main works and conclusions in this dissertation are as follows:(1) Establish the model of forecasting theory line loss rate. The input of the mode is the number and length of transmission line of each voltage level, the number and capacity of transformer of each voltage level, assoeted power sale. The output is theory line loss rate. Make use of BP artificial neural network in MATLAB's toolbox to train and test the sample and achieve the forecasting of line loss rate.(2) Because the model can't compute the amount of excess rising or descending caused by the great change of electrical source distribution and measure device. This dissertation takes into account the finishing item of planning or reconstructing to modify the base quantity of line loss rate. The item includes: long distance transmission, wheeling service, add reactive power compensation, regulating three phase unbalance power of distribution transformer, and so on. Then the result will accord with fact.(3) According to the network data and power supply data provided by A power company, this dissertation forecasts power supply of the power company and 12 presidial power supply bureaus in 2006 first. Second, make use of the BP ANN to forecast the base quantity of line loss rate of 12 power supply bureaus and A power company in 2006. Then modify the base quantity by the amount of some items. Functionary of A power company thinks better of the forecasting method proposed in this dissertation, and considers that the result is helpful to establish the line loss rate.