Research Of Power Consumption Load Forecasting In Jinan

The precise load forecasting must be based on vast, overall and accurate data of electric power system burden and social economic development. It was affected by the factors of economic development, the structure of industry, residents'income level, climate, national policies, etc, so it is a typical gray system. Gray system theory provides effective analysis tool to analyze and dispose leanness information systems to search its evolvement law and set up prediction model by generating operation and relative analysis technique, etc. However, GM(1,1) model is a exponential function which is more applicable to a power consumption's slower growth, but it is very difficult to change strictly according to the index rule, which causes rate of increment excessively quickly and lower forecast precision when carrying on the long-term forecast. Therefore GM (1,1) model has the certain limit in the practical application. Based on this, set up GM(1,1,θ) model, solved the parameters by Ant Colony Algorithm, finally apply neural network to optimize residual error. Case analysis has indicated that the method has the certain practical value.