半参数回归分析法在电力负荷预测中的应用  

Application of Semi-parametric Regression Analysis Method in Electric Power Load Forecasting

作　　者：杨丽徙[1] 张永锋[1] 许向伟 冯越 

机构地区：[1]郑州大学电气工程学院,河南郑州450001 [2]遂平县电业公司,河南遂平463100 [3]驻马店供电公司,河南驻马店463400

出　　处：《郑州大学学报（工学版）》2010年第3期29-32,共4页Journal of Zhengzhou University（Engineering Science）

基　　金：河南省教育厅自然科学基金资助项目(2009A470008)

摘　　要：为了融合参数模型和非参数模型的优点,对半参数回归模型在电力负荷预测中的应用进行了研究.在原始数据预处理的基础上,运用相关性分析理论确定了半参数模型中的参数与非参数变量,构建了半参数回归模型;利用两阶段最小二乘法和权重系数法对回归系数和回归参数进行了估计.算例结果表明,半参数回归分析法具有较高的负荷预测精度和较广泛的适用范围.In order to integrate the advantages of parameter model and non-parameter model,a semi-parametric regression model was studied for power load forecasting.On the basis of pretreatment of the original data,the use of relevant theory determined the parameters of semi-parametric model and non-parametric variables and constructed the semi-parametric regression model.The regression coefficients and regression parameters were assessed with the two-stage least squares method and the weight coefficients.The example shows that this method is more precise and has broader adaptability.

关 键 词：半参数回归模型 两阶段最小二乘法 权重系数法 参数 非参数 

分 类 号：TM715[电气工程—电力系统及自动化]