Design And Implementation Of Short Term Power Load Forecatinging System Based On Grey Theory

In recent years,with the continuous improvement of China's market economy,the power industry is gradually transition to the market process,the accuracy,real-time and reliability of load forecasting have put forward higher request.Due to the power system in each region is different,it is difficult to find a common model to apply all the circumstances.Therefore,this paper based on a project which belong to the Technology Park in Yunnan,proposed a prediction model suitable for the actual situation,and considered the weather factors to improve the prediction accuracy.Before the design of short-term load forecasting,by analyzed the regularity of historical load data of Technology Park in Yunnan,the date type was divided into two types of working days and weekends,solved the problem on choice of the similar day about history load.Focused on the Pre-processing method for data loss and data skip problem in load history data;quantitative processing on the temperature and the humidity of the meteorological factors to establish the method of factor and the improved grey model by Markov.Finally,the design of the short-term power load forecasting system software based on grey theory,and through friendly interface to display the results of the load forecast.Feasibility validation of the model shows that for relatively stable weather conditions in Yunnan and the volatility of the original data is not too large.In this paper,the average of load forecasting accuracy of working day and weekend is 95%,and it is improved by about 5% compared with the common gray forecasting model,with the accuracy of GM(1,n)prediction model is essentially flat,but the algorithm time complexity is low.The test shows the feasibility of the system.