基于动态熵权的短期风速组合预测
|
摘要
为了提高风电场风速预测的准确性,将不同预测方法的权重推广到权重序列,生成权重矩阵,同时采用新的预测误差更新权重矩阵,获得所需模型.建立三种单一预测模型,统计它们十天的预测误差,获得误差序列,在此基础上,提出动态熵权法.采用熵权法确定各单一预测模型在96个预测时刻的权值,并根据新的24小时预测误差更新误差序列和权重矩阵,从而获得动态组合预测模型.结果表明,动态组合预测模型的整体误差指标比单一预测模型较小,预测精度显然增高,证明了所建模型有效且实用.
In order to improve the accuracy of wind speed forecast in wind farm,the weights of different forecast methods were extended to weight series so as to generate the weight matrix. Meanwhile,the newforecast error was used to update the weight matrix,and the required model was obtained. Three single forecast models were established,and the forecast errors of the models in 10 days were counted,and the error series was obtained. On this basis,the dynamic entropy method was proposed. In addition,the weights of each single forecast model at 96 forecast moments were determined with the entropy weight method,and the error series and weight matrix were updated according to new24 h forecast error so as to obtain the dynamic combination forecast model. The results showthat compared with the sigle forecast model,the dynamic combination forecast model has smaller overall error index,and the forecast accuracy gets effectively improved,which proves the validity and practicability of the proposed model.
In order to improve the accuracy of wind speed forecast in wind farm,the weights of different forecast methods were extended to weight series so as to generate the weight matrix. Meanwhile,the newforecast error was used to update the weight matrix,and the required model was obtained. Three single forecast models were established,and the forecast errors of the models in 10 days were counted,and the error series was obtained. On this basis,the dynamic entropy method was proposed. In addition,the weights of each single forecast model at 96 forecast moments were determined with the entropy weight method,and the error series and weight matrix were updated according to new24 h forecast error so as to obtain the dynamic combination forecast model. The results showthat compared with the sigle forecast model,the dynamic combination forecast model has smaller overall error index,and the forecast accuracy gets effectively improved,which proves the validity and practicability of the proposed model.
引文
[1]张国强,张伯明.基于组合预测的风电场风速及风电机功率预测[J].电力系统自动化,2009,33(18):92-95.(ZHANG Guo-qiang,ZHANG Bo-ming.Wind speed and w ind turbine output forecast based on combination method[J].Automation of Electric Pow er Systems,2009,33(18):92-95.)
[2]丁华杰,宋永华,胡泽春,等.基于风电场功率特性的日前风电预测误差概率分布研究[J].中国电机工程学报,2013,33(34):136-144.(DING Hua-jie,SONG Yong-hua,HU Ze-chun,et al.Probability density function of day-ahead w ind pow er forecast errors based on pow er curves of w ind farms[J].Proceedings of the CSEE,2013,33(34):136-144.)
[3]Catalao J P S,Pousinho H M I,Mendes V M F.Short-term w ind pow er forecasting in Portugal by neural netw ork and w avelet transform[J].Renew able Energy,2011,36(4):1245-1251.
[4]冯双磊,王伟胜,刘纯,等.基于物理原理的风电场短期风速预测研究[J].太阳能学报,2011,32(5):611-616.(FENG Shuang-lei,WANG Wei-sheng,LIU Chun,et al.Short-term w ind speed prediction based on physical principle[J].Acta Energiae Solaris Sinica,2011,32(5):611-616.)
[5]蒋小亮,蒋传文,彭明鸿,等.基于时间连续性及季节周期性的风速短期组合预测方法[J].电力系统自动化,2010,34(15):75-79.(JIANG Xiao-liang,JIANG Chuan-wen,PENG Minghong,et al.A short-term combination w ind speed forecasting method considering seasonal periodicity and time-continuity[J].Automation of Electric Pow er Systems,2010,34(15):75-79.)
[6]蔡凯,谭伦农,李春林,等.时间序列与神经网络法相结合的短期风速预测[J].电网技术,2008,32(8):82-85.(CAI Kai,TAN Lun-nong,LI Chun-lin,et al.Shortterm w ind speed forecasting combing time series and neural netw ork method[J].Pow er System Technology,2008,32(8):82-85.)
[7]夏琳琳,台金娟,刘惠敏,等.权重提取与Dempster多重融合的凝汽器真空预测[J].沈阳工业大学学报,2015,37(3):329-334.(XIA Lin-lin,TAI Jin-juan,LIU Hui-min,et al.Condenser vacuum prediction based on w eight extraction and Dempster multiple fusion[J].Journal of Shenyang University of Technology,2015,37(3):329-334.)
[8]杨锡运,刘欢,张彬,等.基于熵权法的光伏输出功率组合预测模型[J].太阳能学报,2014,35(5):744-749.(YANG Xi-yun,LIU Huan,ZHANG Bin,et al.A combination method for photovoltaic pow er forecasting based on entropy w eight method[J].Acta Energiae Solaris Sinica,2014,35(5):744-749.)
[9]刘兴杰,米增强,杨奇逊,等.基于经验模式分解和时间序列分析的风电场风速预测[J].太阳能学报,2010,31(8):1037-1041.(LIU Xing-jie,MI Zeng-qiang,YANG Qi-xun,et al.Wind speed forecasting based on EM D and time-series analysis[J].Acta Energiae Solaris Sinica,2010,31(8):1037-1041.)
[10]蒋金良,林广明.基于ARIMA的自动站风速预测[J].控制理论与应用,2008,25(2):374-376.(JIANG Jin-liang,LIN Guang-ming.Automatic station w ind speed forecasting based on ARIM A model[J].Control Theory&Applications,2008,25(2):374-376.)
[11]王德明,王莉,张广明.基于遗传BP神经网络的短期风速预测模型[J].浙江大学学报(工学报),2012,46(5):837-840.(WANG De-ming,WANG Li,ZHANG Guang-ming.Short-term w ind speed forecast model for w ind farms based on genetic BP neural netw ork[J].Journal of Zhejiang University(Engineering Science),2012,46(5):837-840.)
[12]郭琪.基于NWP的风电负荷预测方法在内蒙古电网中的应用[D].天津:天津大学,2010.(GUO Qi.The application of NWP-BASED wind pow er load prediction method in inner M ongolia pow er grid[D].Tianjin:Tianjin University,2010.)
[13]孙培学,赵坤鹏.基于熵权法的组合模型在大坝渗流预测中的应用[J].水电能源科学,2013,31(12):70-73.(SUN Pei-xue,ZHAO Kun-peng.Application of combination model based on entropy method in dam seepage forecasting[J].Water Resources and Pow er,2013,31(12):70-73.)
[14]欧阳森,石怡理.改进熵权法及其在电能质量评估中的应用[J].电力系统自动化,2013,37(21):156-164.(OUYANG Sen,SHI Yi-li.A new improved entropy method and its application in pow er quality evaluation[J].Automation of Electric Power Systems,2013,37(21):156-164.)
[2]丁华杰,宋永华,胡泽春,等.基于风电场功率特性的日前风电预测误差概率分布研究[J].中国电机工程学报,2013,33(34):136-144.(DING Hua-jie,SONG Yong-hua,HU Ze-chun,et al.Probability density function of day-ahead w ind pow er forecast errors based on pow er curves of w ind farms[J].Proceedings of the CSEE,2013,33(34):136-144.)
[3]Catalao J P S,Pousinho H M I,Mendes V M F.Short-term w ind pow er forecasting in Portugal by neural netw ork and w avelet transform[J].Renew able Energy,2011,36(4):1245-1251.
[4]冯双磊,王伟胜,刘纯,等.基于物理原理的风电场短期风速预测研究[J].太阳能学报,2011,32(5):611-616.(FENG Shuang-lei,WANG Wei-sheng,LIU Chun,et al.Short-term w ind speed prediction based on physical principle[J].Acta Energiae Solaris Sinica,2011,32(5):611-616.)
[5]蒋小亮,蒋传文,彭明鸿,等.基于时间连续性及季节周期性的风速短期组合预测方法[J].电力系统自动化,2010,34(15):75-79.(JIANG Xiao-liang,JIANG Chuan-wen,PENG Minghong,et al.A short-term combination w ind speed forecasting method considering seasonal periodicity and time-continuity[J].Automation of Electric Pow er Systems,2010,34(15):75-79.)
[6]蔡凯,谭伦农,李春林,等.时间序列与神经网络法相结合的短期风速预测[J].电网技术,2008,32(8):82-85.(CAI Kai,TAN Lun-nong,LI Chun-lin,et al.Shortterm w ind speed forecasting combing time series and neural netw ork method[J].Pow er System Technology,2008,32(8):82-85.)
[7]夏琳琳,台金娟,刘惠敏,等.权重提取与Dempster多重融合的凝汽器真空预测[J].沈阳工业大学学报,2015,37(3):329-334.(XIA Lin-lin,TAI Jin-juan,LIU Hui-min,et al.Condenser vacuum prediction based on w eight extraction and Dempster multiple fusion[J].Journal of Shenyang University of Technology,2015,37(3):329-334.)
[8]杨锡运,刘欢,张彬,等.基于熵权法的光伏输出功率组合预测模型[J].太阳能学报,2014,35(5):744-749.(YANG Xi-yun,LIU Huan,ZHANG Bin,et al.A combination method for photovoltaic pow er forecasting based on entropy w eight method[J].Acta Energiae Solaris Sinica,2014,35(5):744-749.)
[9]刘兴杰,米增强,杨奇逊,等.基于经验模式分解和时间序列分析的风电场风速预测[J].太阳能学报,2010,31(8):1037-1041.(LIU Xing-jie,MI Zeng-qiang,YANG Qi-xun,et al.Wind speed forecasting based on EM D and time-series analysis[J].Acta Energiae Solaris Sinica,2010,31(8):1037-1041.)
[10]蒋金良,林广明.基于ARIMA的自动站风速预测[J].控制理论与应用,2008,25(2):374-376.(JIANG Jin-liang,LIN Guang-ming.Automatic station w ind speed forecasting based on ARIM A model[J].Control Theory&Applications,2008,25(2):374-376.)
[11]王德明,王莉,张广明.基于遗传BP神经网络的短期风速预测模型[J].浙江大学学报(工学报),2012,46(5):837-840.(WANG De-ming,WANG Li,ZHANG Guang-ming.Short-term w ind speed forecast model for w ind farms based on genetic BP neural netw ork[J].Journal of Zhejiang University(Engineering Science),2012,46(5):837-840.)
[12]郭琪.基于NWP的风电负荷预测方法在内蒙古电网中的应用[D].天津:天津大学,2010.(GUO Qi.The application of NWP-BASED wind pow er load prediction method in inner M ongolia pow er grid[D].Tianjin:Tianjin University,2010.)
[13]孙培学,赵坤鹏.基于熵权法的组合模型在大坝渗流预测中的应用[J].水电能源科学,2013,31(12):70-73.(SUN Pei-xue,ZHAO Kun-peng.Application of combination model based on entropy method in dam seepage forecasting[J].Water Resources and Pow er,2013,31(12):70-73.)
[14]欧阳森,石怡理.改进熵权法及其在电能质量评估中的应用[J].电力系统自动化,2013,37(21):156-164.(OUYANG Sen,SHI Yi-li.A new improved entropy method and its application in pow er quality evaluation[J].Automation of Electric Power Systems,2013,37(21):156-164.)