Residential Electricity Demand Analysis And Forecasting Model

With the development of economy and adjustment of economic structure in China, the proportion of the residential electricity consume in total electricity consume is increasing gradually. Therefore, it is necessary to investigate the model of residential electricity demand for accommodating to the rapid increase of residential electricity demand. The paper summarizes the advantages and disadvantages of the past residential electricity demand models. After analyzing the economic and non-economic factors which effect the residential electricity demand, the paper puts forward models more suitable for market economy and gets more satisfied result. Based on the traditional models, the paper contributes to the following aspects: 1. A time-series linear expenditure system model is introduced for residential electricity demand forecast. And a cross-section expenditure model is used to calculate the income elasticity and price elasticity. 2. By setting up dummy variable regression equation, the effect of non-economy factors is analyzed quantitatively. 3. Using artificial neural network to map the relationship between the household appliance percentage and average residential electricity demand, this paper presents an artificial intelligent method to forecast residential electricity demand. The results of a practical simulation illustrate the accuracy and validity of this method. 4. Since the social behavior is important in the long term residential electricity demand, the paper discusses the human factors in the residential electricity demand in the long run, as a basis for the future study.