基于灰色关联度与LSSVM组合的月度负荷预测.pdfVIP

第 36 卷 第 8 期 电 网 技 术 Vol. 36 No. 8 2012 年 8 月 Power System Technology Aug. 2012 文章编号：1000-3673 （2012 ）08-0228-05 中图分类号：TM 714 文献标志码：A 学科代码：470·4054 基于灰色关联度与 LSSVM 组合的月度负荷预测 1 1 1 2 刘文颖 ，门德月 ，梁纪峰 ，王维洲 （1．新能源电力系统国家重点实验室(华北电力大学) ，北京市 昌平区 102206； 2．甘肃电力科学研究院，甘肃省 兰州市 730050 ） Monthly Load Forecasting Based on Grey Relational Degree and Least Squares Support Vector Machine 1 1 1 2 LIU Wenying , MEN Deyue , LIANG Jifeng , WANG Weizhou (1. State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources (North China Electric Power University), Changping District, Beijing 102206, China; 2. Electric Power Research Institute of Gansu Province, Lanzhou 730050, Gansu Province, China) ABSTRACT: Due to the dual trends of monthly load, i.e., the 摘要：由于月度负荷的二重趋势特性，其变化呈现出复杂的 increasing and fluctuation, its variation is indicative of the 非线性组合特征，使预测精度一直不能达到令人满意的结 feature of complex nonlinear combination, so its forecasting 果。针对月负荷的二重趋势特性和最小二乘支持向量机 accuracy cannot be satisfactory. In allusion to the duplicity of (least squares support vector machine，LSSVM)存在的数据输 monthly load and in view of such defects in least squares 入维数大、训练时间长等缺点，提出一种基于灰色关联度与 support vector machine (LSSVM) as high dimensionality and LSSVM 组合的月度负荷预测方法。该方法通过计算灰色关 long training period, combining grey relational degree