Load Forecasting Analysis And Management In One Of The Study And Practice

Load forecast is one of the important work for electric power dispatching, trading and planning in electric power company. Load forecasting accurately which is the foundation of exact operation scheduling to keep the power grid operated in the secure and reliable state. Load forecasting is also the key of making trading between generation units, which can guarantee the well-balanced operation of electricity market. At the same time, it is the basis for demand side management and electric power planning. With the development and reconstructure in the power industry, it is urgent to improve the load forecasting accuracy.According to the current load forecasting models in application, the concept of standardization for load-related factors, bad data t-testing and periodical load forecasting is introduced, which improves the load forecasting result.Integrated forecasting model and algorithm, which combines the virtue of each model, are designed to achieve more accurate forecasting result in this thesis, the accuracy of integrated model is more steady-going than those basic single models. It optimizes those forecasting results of several basic models according to the idea of Simulative Forecasting considering the impacts of load-related factors, so the integrated model can get a much better and steady accuracy than most basic forecasting models. The novel idea of integration of power grid load forecast, analysis and management is proposed in this thesis, for the existing simple and inefficient function of load forecast software applied in china. Via mutual revision and comprehensive load characteristic analysis from the provincial and regional dispatching units, it can make up the deficiency of load forecasting model.