基于时变参数模型的中长期负荷预测  

Medium and long-term load forecasting based on time-varying parameter model

作　　者：李翔[1] 陈昊[2] 

机构地区：[1]国电南瑞科技股份有限公司电网控制分公司,南京210061 [2]南京供电公司,南京210008

出　　处：《电力需求侧管理》2009年第2期32-34,共3页Power Demand Side Management

摘　　要：提出了一种基于状态空间时变参数模型的中长期负荷预测建模新方法,将状态变量纳入可观模型求解,反映了变量间均衡关系的变化规律。实证部分使用中国年度用电量与国内生产总值时间序列建立了模型,对2002—2005年度的用电量进行了测算,检验了方法的有效性。A new method of medium and long term load forecast model based on state space time-varying parameter model is presented. The state variable is brought into observable model tor solution, which reflects the varying rules of equilibria relation of variables. In demonstration section, the model with China yearly electricity consumption and GDP is estimated, and the forecast pertformance of 2002--2005 yearly electricity consumption is provided. The efficiency of this method is validated.

关 键 词：负荷预测 卡尔曼滤波 状态空间 时变参数模型 

分 类 号：F407.61[经济管理—产业经济]