The Research And Application Of The Uncertain Models For Medium And Long Term Load Forecasting

Medium and long term load forecasting is the basis of power system planning and construction. Exact degree of load forecasting will directly affect the rationality of investment, network layout and running. For the complexity and uncertainty of load, the uncertain models for medium and long term load forecasting are discussed.The significance of the research is given. According to the characters of the uncertain factors, this paper sets forth the feasibility and the advantage of the stochastic theory, fuzzy theory, gray system principal, cloud theory, blind number theory and the theorem of unascertained rational number in the aspect of solving the problem of the uncertain factors.In the medium and long term load forecasting, the shortage and damage data may be encountered when historical data are collected. According to that, this paper applies a number of methods including two kinds of neural network methods to data pretreatment for electric load forecasting. The complement of the vacant data and the searching and correcting of the distortional data are discussed. Besides, the disposal of limiting power by turning off pull-switch is explained too.The central problem of power system load forecast is the proper use of certain mathematical method. Because of the influence of the uncertain factors to the medium and long term load forecasting, this paper applies some new methods to improve the original models, or to presents some new models, such as random predicting model, fuzzy forecasting model, blind number forecasting model, cloud forecasting model and unascertained rational number combinatorial predication model and so on. The examples tell that these models are more applicable for the medium and long term load forecasting.On the basis of the research above and combining the actual requirement of an innovatory project, a set of load forecasting software with Delphi being make up of conventional methods and uncertain methods was developed.In the end, the research results of this dissertation are summarized, and some directions for further research are also provided.