Research On Regional Long Term Power Load Forecasting Model And Its Application

Power system load forecasting is an important work of the power sector, is thepremise and foundation of economic operation and the power system planning,decision-making, the accuracy of the load forecast is of great significance to thedevelopment of the national economy and the safety economic operation of powersystems in regions. According to the general situation of West Baotou area, and thedevelopment of social economy in "eleven five Year Plan" period, under regionalpower load history, analyze regional power grid load forecasting in” Twelfth FiveYear Plan" period.There are many factors affecting the power load, but the industrial structure hasan influence on demand of electricity, which can not be ignored. Industrial structureoccupies the dominant position in the national economy, and the electric powerindustry is not only the premise of the national economy development, but also thebasis for social development and people’s living standards improved. Relationshipbetween electric power demand and the industrial structure, the development of thenational economy, lead to the change of industrial structure, which caused changes inall sectors to power demand, and then influence the development on electric powerindustry.Considering the influence of industrial structure on the power load, based onthe analysis of the relationship between the West Baotou area industrial structure andpower requirements, this paper combined with the power load prediction model, hasestablished the long-term power load prediction model of industrial structure inBaotou city. This paper applies support load vector machine and grey predictionmodel to predict the west of Baotou City, compared two methods of forecastingresults and analyze, to draw more accurate power load forecasting method of the area.The results show that the prediction accuracy test. The results of the two methods arequite good, but the accuracy of support vector machine prediction method is high, thepredicted result is more reliable, and more suitable for medium and long term powerload forecasting.