短期母线负荷优性组合预测方法及其在成都电网中的应用

Research on Superior Combined Forecasting Method of Short-Term Bus Load and Its Application in Chengdu Grid

短期母线负荷预测是编制停电检修计划、确定年度运行方式、优化改接负荷及状态估计等工作的重要前提,对提高驾驭大电网能力、开展节能发电调度、实现安全校核精益化管理具有重要意义。针对短期母线负荷预测中单一预测方法的局限性,提出了优性组合预测新方法。依据预测方法在样本区间及预测区间上的灰色关联度、点预测精度及区间预测有效度指标,提出了预测方法综合有效度系列概念。以全区间综合有效度为优选指标,对预测方法进行筛选,采用方差-协方差法求取组合权重。成都电网110 kV母线负荷预测算例表明,所提方法预测精度高、稳定性好。

Short-term bus load forecasting is the important premise for formulating overhaul plan, determining the annual operation mode and optimizing the load change and state estimation, and is very important for energy-saving generation dispatching and lean management of safety operation. In order to solve the shortcoming of the single forecasting method, a new superior combined forecasting method is proposed in this paper. According to the gray correlation index, point prediction accuracy index and interval forecasting effectiveness index in the sample interval and prediction interval, a series of concepts of comprehensive effectiveness of the forecasting method are introduced. By using the variance and covariance method, the weight coefficient of corresponding model is calculated. A case study of the 110 kV bus load in Chengdu grid shows that the proposed combined method is effective with high accuracy and stability.