多维度饱和负荷预测方法及其应用  

Forecasting Method and Application of Multi-dimensional Saturated Load

作　　者：刘杰锋[1] 程浩忠[1] 韩新阳[2] 尤钟晓[1] 杨宗麟[3] 

机构地区：[1]上海交通大学电力传输与功率变换控制教育部重点实验室,上海200240 [2]国网能源研究院,北京100052 [3]华东电网有限公司,上海200002

出　　处：《电力系统及其自动化学报》2015年第2期44-50,共7页Proceedings of the CSU-EPSA

基　　金：2012年度上海市优秀学术带头人计划资助项目(12XD1402900)

摘　　要：电力饱和负荷预测可以确定电网的最终规模,指导电网的改造与建设,有效降低其成本。也可以更好地指导变电站布点、线路走廊的规划和建设工作。多维度饱和负荷预测方法首先通过灰色关联度分析评估影响因素,然后选择建立以影响因素为自变量、用电需求为因变量的函数,通过最小二乘法确定具体的数学模型。该方法不仅可以评估未来用电的饱和情况,也可以对影响因素进行灵敏度分析,具有比较高的预测精度,是对饱和负荷预测方法的拓展与创新。自身也可以产生反馈校验指标,具有一定的实用性与指导性价值。提出的饱和负荷判定指标验证了最终预测结果的正确性与合理性。Saturated power load forecasting can determine the ultimate size of the grid, guide the renovational and constructional work of power grid and reduce its cost more effectively, as well as guiding the layout of the transformer substations, the line corridor planning and construction. In this paper, multi-dimensional saturated load forecasting method is adopted, which firstly evaluates the influential factors through the grey correlation degree analysis, and then chooses the influential factors as independent variables, and the electricity demand as the dependent variable to create function, and the specific mathematical model can be determined by least squares method. This method, with rela- tively high prediction accuracy, can not only assess the future electricity saturated conditions, but also analyze the sen- sitivity to the influential factors, that is the development and innovation to the saturated forecasting method. This method can also generate feedback parity index, with a certain degree of practicality and instructive significance. Satu- rated load indicators are proposed to validate the correctness and rationality of the final predicted results.

关 键 词：饱和负荷 灰色关联度 最小二乘法 影响因素 多维度 

分 类 号：TM715[电气工程—电力系统及自动化]