Power Load Forecasting Based On Urban Economy Theory

City is the center of human society, economy, and cultural activities. From the aspect of the development situation of the whole world, the faster the economy growth, the higher the material culture level is, then the higher the degree of dependence on electricity and the consumption level of electricity are higher. With the increasing development of urban economy, the urban electricity will own corresponding characteristics. So, it has the theoretic and realistic significance to study how to forecast the urban power demand through considering the development of urban economy.In this paper, at first, through analyzing on the factors that affect economic growth of the typical cities in the world, such as the urban population, the industrial structure, the urbanization rate, the urban land, the urban strategy and the policy, the urban environment, and so on,the laws of urban economic development is researched. And the saturated situation of future urban economic development is also analyzed. The evaluation index, which is used to judge whether urban economic development is saturated, is established.Then,based on the risk element transmission theory, the relationship between urban economy and urban power load forecasting is studied; the result shows that gross domestic product (GDP), urbanization rate and proportion of added value of tertiary industry have higher relevance with the fluctuation of electricity. Finally, the growth rate method, the electricity consumption per capita method, the Logistic curve method, the exponential smoothing method and other load forecasting methods are introduced in this paper. But the precisions of forecasting results are low in these method. So, an urban load forecasting model is established based on the risk element transmission theory through considering the risk factors of urban economy. In order to ensure the accuracy, a combination method based on the ant colony algorithm is put forward. By this method, the objective function is formulated according to the least squares criterion in order to obtain the weighted coefficients; then the single forecasting model is selected by the weighted coefficients. In addition, taking Beijing city as an example, the empirical analysis results show that this model is scientific and has certain application values.