基于DPBIL-SVM混合模型的电力现货市场出清价预测研究  

Power Spot Market Clearing Price Forecasting Based on DPBIL-SVM Hybrid Model

作　　者：张帅 陈仕军[1,2] 马光文[1,2] 黄炜斌[1,2] 陶春华 ZHANG Shuai;CHEN Shi-jun;MA Guang-wen;HUANG Wei-bin;TAO Chun-hua(Collegeg of Water Resources and Hydropower,Sichuan University,Chengdu 610065,China;State Key Laboratory of Hydraulics and Mountain River Engineering,Sichuan University,Chengdu 610065,China;Dadu River Hydropower Development Co.,LTD,Chengdu 610041,China)

机构地区：[1]四川大学水利水电学院,四川成都610065 [2]四川大学水力学与山区河流开发保护国家重点实验室,四川成都610065 [3]国家能源集团大渡河流域水电开发有限公司,四川成都610041

出　　处：《水电能源科学》2020年第4期197-200,共4页Water Resources and Power

基　　金：国家重点研发计划(2018YFB0905204);四川大学专职博士后研发基金(2018SCU12062)。

摘　　要：为准确预测现货市场出清价,利用改进的基于种群增量学习的进化算法(DPBIL)对SVM参数进行优化,构建了基于DPBIL-SVM的混合预测模型,将该模型应用于挪威电力市场短期电价预测中,并与灰色GM(1,1)模型和BP人工神经网络模型进行比较。结果表明,所提模型能够将现货市场出清价预测误差控制在5%以下,合格率97%,效果优于灰色GM(1,1)模型和BP人工神经网络模型,符合现货市场实际报价的要求。In order to predict the clearing price of spot market accurately,the improved evolutionary algorithm based on population learning was used to optimize the SVM parameters,and a hybrid prediction model was constructed based on the DPBIL-SVM.The model was applied to the short-term electricity price prediction of Norwegian electricity market.Compared with the grey GM(1,1)model and BP artificial neural network model,the results show that the proposed model can control the clearing price prediction error of the spot market to less than 5%,and the qualified rate was 97%.The prediction effect was better than that of the GM(1,1)and BP artificial neural network model,which meets the actual quotation requirements of the spot market.

关 键 词：电力市场 出清价预测 PBIL SVM 人工智能 

分 类 号：TM715[电气工程—电力系统及自动化] F407.6[经济管理—产业经济]