Research On City Saturated Power Load Forecasting Based On Combined Logistic Model

Power load presented different features in different stages of economic and social development.When economic and society developed to a certain stage,power load growth would slow and even stopped growing under the restrictions,such as regional energy structure,land resources,environment,population,urban power grid planning and policy factors.The power land tended to saturation.Saturated load,which is the key indicator to determine the power grid scale in urban power grid planning development,can determine the final size of urban electric power development so as to make room for the development of urban power in the future.It also can provide basic data and guide the construction and renovation of urban power grid for the power supply,the grid of the long-term planning and distribution network.This article analysis the development rule of city load among the typical developed countries and summarizes the development of power load and the saturation stage of social and economic characteristics.Based on the above research,improve the electrical load on quantitative index system into the saturation stage of development and introduces the commonly used method of saturated load forecasting model and scope of application.The consciousness model of power load factors is established and the hierarchical relationship between influence factors is analyzed by the explanation structure.It's concluded that the factors can be divided into surface reasons,shallow reasons and deep reasons.Considering the characteristics of traditional logistic model,in the view of method accumulated generating data by Grey theory and the grade difference format,the poor gray logistic model is set up.It avoids the subjective parameters in traditional forecasting model.Then,considering the factors that has a direct impact on power load in the surface layer,using PSO to optimize the neural network model to function the parameter that is the growth speed in the logistic model,the logistic development model is established.Through the calculation of variance-covariance and degree of deviation of single model to determine the weight of combination forecast model.The model can avoid the single forecasting shortcomings,enhance the practicability and flexibility and improve the prediction precision.Finally the article analyzes the Beijing economy,population and electric power development situation.The Beijing future electricity demand is forecasted by the historical data.The results show that the electricity demand is less than 2% a year in Beijing from2026-2027,the total electricity demand is among 1331.68-135.091 billion KWH,the city enters the saturation stage of development.