Short-term Power Load Forecasting Study

Load forecast is the foundation of power system operating dependably and economically. The result of load forecasting has become on important basis for operating management, economic dispatch, increasing system's capacity, and pursuing power market. In view of the network operating conditions in an area, this paper analyzes the present situation of the short-range electric load forecasting and mathematical models of various forecasting, proposes a new mathematical model, which conforms to the electric network load of this area, and also develops an appropriate application software for short-range electric load forecasting. Because of the randomness, the periodicity and the impact property of load changes, the short-range electric load forecasting accuracy is related to the pre-processing of original data, the load forecasting model, the sudden change of climate etc. Without an appropriate mathematical model, it is difficult to meet the demand of farecasting accuracy by using computer software. By combining the advanced modern mathematical modeling theory with the advanced database application and development tools and software engineering, this paper proposes the new mathematical model, and through programming realizes the 24-hour data forecasting of punctual load, daily peak-to-valley load and daily average load in the area. This system adopts cumulatively autoregressive moving average model [ARIMA] of time series method and modified model GM( 1,1) of grey system, makes a local load forecasting modeling through the integration of the above two models and also preprocesses the daily load during the sudden change of climate, thus greatly improving the forecast accuracy. The practical operation indicates that the model is reasonable and easy to operate with complete function. It also shows that the software interface is friendly and the forecast accuracy is up to the standard, which meets the demand of the consumers. This system improves the operating management of power system to a higher level, and has a better practical application value.