基于气象综合指数和加权LSSVM的短期电力负荷组合预测  

Short-term electric load combination forecasting based on meteorological composite index and weighted LSSVM

作　　者：吕何 孔政敏 陈培垠 刘晓帆 LV He;KONG Zhengmin;CHEN Peiyin;LIU Xiaofan(School of Electrical Engineering,Wuhan University,Wuhan 430072,China)

机构地区：[1]武汉大学电气工程学院

出　　处：《电气应用》2019年第10期42-50,共9页Electrotechnical Application

基　　金：国家自然科学基金（51707135）

摘　　要：短期电力负荷易受气象情况、日期属性和偶然因素等影响,预测准确度较低。为解决上述问题,提出了基于气象综合指数和加权最小二乘支持向量机(WLS-SVM)的电力负荷组合预测方法。分析气象因子耦合作用引入气象综合指数进行短期负荷预测。采用线性映射与分区映射方法处理负荷特征,解决了数据之间的差异性,并采用灰色关联分析得到气象综合指数与电力负荷的关联度,其权重系数由关联度确定。在优化最小二乘支持向量机(LSSVM)上将径向基核函数(RBF)和权重系数相结合得到WLSSVM。最后提出滚动窗口预测法并建立短期负荷组合优化预测模型,降低偶然误差对负荷预测结果的影响。以我国南方某电网公司每日96点负荷历史数据为实例样本进行仿真,结果表明所提出方法与LSSVM预测方法相比预测准确度更高,逐时预测性能优于逐天预测,验证了所提方法的有效性。Easily affected by meteorological information, date attributes, accidental factors, etc., the accuracy of short-term load forecasting is generally low. In order to solve the above problems, A combination forecasting method based on meteorological comprehensive index and weighted least squares support vector machine to forecast short-term load is proposed. Meteorological comprehensive index for short-term load forecasting is introduced in analyzing the coupling of meteorological factors. The linear mapping and partition mapping methods are used to deal with the load characteristics, and the difference between the data is solved. Grey correlation analysis is used to obtain the correlation between meteorological comprehensive index and load, by which the weight coefficient is determined. A weighted least squares support vector machine(WLS-SVM) is deduced by combining radial basis kernel function(RBF) and the weighting coefficient to optimize least squares support vector machine(LSSVM). Finally, the scrolling window forecast method is proposed to establish the short-term load combination optimization forecasting model to reduce the impact of accidental error on short-term load forecasting. Taking the 96-point load historical data of a power grid in China Southern Grid as a practical example, simulation results show that the proposed method has higher forecasting accuracy than the LSSVM method and the hour-ahead performance of the proposed method is better than dayahead performance, which verifies the effectiveness of the proposed method.

关 键 词：负荷预测 气象综合指数 关联分析 滚动窗口预测法 最小二乘支持向量机 

分 类 号：TM715[电气工程—电力系统及自动化]