极坐标系下的电力负荷混沌局域预测算法  

A Chaotic Local Power Load Forecasting Algorithm in Polar Coordinate System

作　　者：王桓[1] 何怡刚[1] 高坤[1] 

机构地区：[1]湖南大学电气与信息工程学院,湖南省长沙市410082

出　　处：《电网技术》2010年第7期138-142,共5页Power System Technology

基　　金：国家自然科学基金项目(50677014);国家863高技术基金项目(2006AA04A104);新世纪优秀人才支持计划项目(NCET-04-0767)~~

摘　　要：在混沌局域预测算法中,通常使用欧氏距离和关联度衡量相点间的相关性,以确定参考邻域,但这些参量不能有效衡量相点相关性。在极坐标系下,将相点视为向量,则相点间相关性应由向量间的模比值和向量夹角共同确定。根据以上思路,将相点向量的模比值和夹角余弦值融合为一个衡量相点相关性的参量,以确定参考邻域。在预测参数识别中,以相点间的模和夹角作为优化目标,提出了新的线性拟合参数求解算法。将以上算法应用到某城市的电力负荷短期预测中,获得了令人满意的预测效果。In chaotic local forecasting algorithm, Euclid distance and correlation degree are usually used to measure correlativity between phase points to decide the reference neighborhood, however the correlativity of phase points cannot be effectively measured by these parameters. In polar coordinate system the phase points are regarded as vector, in that way the correlativity between phase points should be jointly determined by the ratio of modules between vectors and the included angle of vectors to decide reference neighborhood. According to above idea, the ratio of modules between vectors and the cosine of included angle of vectors are amalgamated into a parameter to measure the correlativity between phase points to decide reference neighborhood. In forecasting parameter identification, the modules and included angles between phase points are taken as optimization objective, then a new algorithm to solve linear fitting parameter is proposed. Applying the proposed algorithm to short-term urban load forecasting, satisfied forecasting results are obtained.

关 键 词：短期负荷预测 混沌 极坐标系 向量 相关性 

分 类 号：TM715[电气工程—电力系统及自动化]