Super Short-term Load Forecast In Power Plant Side

Load forecasting system is based on the natural and capacity increase decision making, social impact, and many other factors, meet certain accuracy under the conditions required, determiNEThe load data for a particular time in the future, load refers to the demand for power or consumption; It is in economic dispatch of power system and an important element of economic operation of power plants.Power load prediction method of predicting prediction method are divided into classic and modern. Classical forecasting methods including: extrapolation of trends, time series analysis, regression analysis, and so on.20century80's later, some modern prediction methods based on emerging theory gradually has been successfully applied. This mainly grey mathematical theory, expert systems, neural networks, fuzzy theory. But the method of application, General focus on the network side, on data access, need to get research targeted network of real-time load generator and key trends of grid nodes. For power generation enterprises, is difficult to timely and accurate access to these data.This study aims to find a case information is not completely symmetrical, this article based on linear extrapolation, full use has access to real-time load data, starting and stopping data generation scheduling and unit, an amendment to the forecast results, elimination of accidents due to sudden load caused by deviation of the predicted value and the actual value. The methods while maintaining speed, improving the overall prediction accuracy of forecast accuracy and inflection points. As far as the station side, this forecast for preventive control, generators of economic operation, enhancing the economic generator, reducing coal consumption, is of great significance.Through this study on a power plant in the period compared to the station data and forecast data in real time, proved the prediction algorithms relative to the traditional advantages of time series algorithms and feasibility in practice. And through Microsoft Visual Studio2005, automatic calculation of super short-term load forecasting system.