The Establishment Of General Forecasting Models And Empirical Research Based On Electricity Market

In the view of common forecasting question in the electric power market, this paper establishes new integrated and combined forecasting models based on various forecasting technologies and methods, and makes use of these improved models to carry through concrete applied research.Among constructing the electricity power demand forecasting model, this paper comprehensive summarizes the characteristics of year electricity consumption, and extracts main characteristics based on typical regressive curve, and carries on residual improvement by the ARMA model, finally integrated typical curve with ARMA to construct new combined forecasting model. At the same time, this paper utilizes the improved model based on exponential regressive curve and ARMA model to carry on the medium and long-term electric power demand forecasting in Hunan province. Moreover, based on qualitative analysis for the electric power demand influencing factors in Hunan province, this paper uses PLS method to carry on quantitative analysis for the electric power demand influencing factors in Hunan province, and brings forward the medium and long-term electric demand forecasting model based on PLS method.Among constructing electricity power load forecasting model, this paper comprehensive summarizes the traits and the influencing factors of electric load. In the view of periodic electric load with double trends of fluctuating and increasing, this paper firstly utilizes ARIAM cumulative model to extract linear characteristic, and secondly makes use of improved GM(1,1) model to extract nonlinear characteristic, finally integrates the two aspects to construct ARIMA-GM integrated forecasting model. This paper makes use of the ARIMA-GM model to carry on short term load forecasting.Among constructing electricity price forecasting model, this paper comprehensive summarizes the traits and the influencing factors of electricity price. In the view of electricity price with periodic and volatility clusters characteristic, this paper firstly utilizes general Auto-regressive Distributed Lag model to describe time characteristic of electricity price series, and secondly obtains simple ARMA model according to test statistics analysis. Further this paper carries on the GARCH effect examination to the ARMA model, and establishes the GARCH model for the residual series of ARMA...