一种计及横向误差校正的短期负荷预测方法

A method of load Forecasting tanking account of transverse error correcting

负荷预报模型由于只能对负荷规律性进行有限描述而在建模域中存在建模误差,在预报域中,与建模误差具有相同统计特性的误差分量就会延续到预报误差中。将频谱分析预测方法与基于灰色理论横向误差校正方法相结合,由频谱分析预测方法进行纵向预测,利用预测时刻之前的整点时刻负荷预测的误差来建立灰色理论模型进行负荷预测的横向误差校正,并对灰色校正模型分点的判断方法作了改进,所提方案在短期负荷预测中的性能良好。

Load forecasting model only can carry on the limited description to the load regularity in the modular field and lead to have the modeling error, so there will be the same statistical property error component with the modeling error to be able to continue to the forecast error in the forecasting field. This paper puts forward the associative method of spectrum analysis and transverse error correcting based on a new gray theory. Therefore, the method of spectrum analysis should be used to forecast lengthways load. We first forecast the load, then transversely correct the error by using the sharp-time forecast error before forecasting time to set up a gray theory model. It improves the point which lies in the method of gray theory correcting. It is shown that the presented scheme possesses good performance for short-term load forecasting.