Research On Application Of Data Mining In Load Forecasting System For Rural Power Supply Enterprise

Power load is the consumption of electricity and electricity in a region. The rural power grid load forecasting is starting from the demand of social economic development in rural areas, and the power system known, considering the rural economy, climate and special events and many other factors, through the analysis and study of rural history data, explore the relationship between the parameters and development of the law in the rural power system, with the economy and climate prediction as the basis of the estimates and forecasts of future rural power demand. The rural power load forecasting is an important part of rural power management system, the future load data provided by it is very important in the controlling, running and planning of the power system.. Due to the old equipment, weak infrastructure, the stable operation of rural power system requires load forecasting in advance, in order to prepare early, make reasonable equipment resources allocation, to ensure safe reliable use in the rural areas.Considering the regional characteristics of power load, this paper analyzes the characteristics of load growth from three aspects of annual load, monthly load and daily load, and analyzes the meteorological sensitivity and major holidays of the load, obtains the general characteristics of rural power network load as a result.Based on the analysis of the power load,we forecast the recent load of Zongyang county by ponit to point ratio method, linear regression method and BP neural network model.we compare the accuracy of the three models to find the best one.The result is positve to State Grid for prepairing the Summer Peak or the Winter Peak.