基于NBA-SVR的日最大负荷预测  

Daily Maximum Load Forecasting Based on NBA-SVR

作　　者：成贵学 陈昱吉 赵晋斌 费敏锐[3] CHENG Gui-xue;CHEN Yu-ji;ZHAO Jin-bin;FEI Min-rui(School of Computer Science and Technology,Shanghai University of Electric Power,Shanghai 200090,China;School of Electrical Engineering,Shanghai University of Electric Power,Shanghai 200090,China;School of Mechanical Engineering and Automation,Shanghai University,Shanghai 200072,China)

机构地区：[1]上海电力大学计算机科学与技术学院,上海200090 [2]上海电力大学电气工程学院,上海200090 [3]上海大学机电工程与自动化学院,上海200072

出　　处：《电工电气》2021年第1期11-16,共6页Electrotechnics Electric

摘　　要：为进一步提高日最大负荷预测精度,提出一种基于新型蝙蝠算法和支持向量回归的日最大负荷预测方法,引入对回波中多普勒效应进行自适应补偿和栖息地选择的新型蝙蝠算法优化选取支持向量回归参数,采用电工杯数学建模竞赛提供的数据训练并建立NBA-SVR模型进行日最大负荷预测,结果表明NBA-SVR模型在预测精度上比BPNN、PSO-SVR、WOA-SVR模型有显著的提升。In order to further improve the accuracy of daily maximum load forecasting,this paper proposed a new daily maximum load forecasting method based on novel bat algorithm optimization and support vector regression.It introduced the adaptive compensation of Doppler effect in the echo and new bat algorithm for habitat selection to optimize the selection of support vector regression parameters.The data provided by the Electrician Mathematical Contest in Modeling are used to train and establish the NBA-SVR model to perform daily maximum load forecasting.The results showed that the NBA-SVR model has better prediction accuracy than the back propagation neural network,PSO-SVR,and WOA-SVR.

关 键 词：日最大负荷预测 新型蝙蝠算法 支持向量回归 参数优化 

分 类 号：TM715[电气工程—电力系统及自动化] TP181[自动化与计算机技术—控制理论与控制工程]