基于多模型融合Stacking集成学习方式的负荷预测方法
|
摘要
人工智能及机器学习技术的快速发展,为负荷预测问题提供了崭新的解决思路。该文结合人工智能的前沿理论研究,提出一种基于多模型融合Stacking集成学习方式的负荷预测方法。考虑不同算法的数据观测与训练原理差异,充分发挥各个模型优势,构建多个机器学习算法嵌入的Stacking集成学习的负荷预测模型,模型的基学习器包含XGBoost树集成算法和长短记忆网络算法。算例使用ENTSO中瑞士负荷数据对算法有效性进行了验证。预测结果表明,XGBoost、梯度决策树、随机森林模型能够通过自身模型的增益情况对输入数据的特征贡献度进行量化分析;Stacking中各个基学习器的学习能力越强,关联程度越低,模型预测效果越好;与传统单模型预测相比,基于多模型融合的Stacking集成学习方式的负荷预测方法有着较高的预测精度。
The rapid development of artificial intelligence and machine learning technology provides an innovative solution for load forecasting. A load forecasting method based on multi-model combination under Stacking framework was proposed, associated with the frontier theory research of artificial intelligence. Considering the difference of data observation and training principles, the Stacking based load forecasting model embedded various machine learning algorithms was proposed to utilize their diversified strength.The XGBoost algorithm constructed by tree ensemble model and long-short term memory were involved in Stacking base-learner layer. Swiss load data in ENTSO was used to verify the feasibility of the algorithm in case study. The forecasting results show that the XGBoost, the gradient decision tree and the random forest model can quantify the contribution of the input data through the gain of their own model. The load forecasting results are more accurate when each of base-learner has lower correlation coefficient in Stacking. The results indicate the Stacking ensemble learning based on multi-model has better prediction performance compared with the traditional single model.
The rapid development of artificial intelligence and machine learning technology provides an innovative solution for load forecasting. A load forecasting method based on multi-model combination under Stacking framework was proposed, associated with the frontier theory research of artificial intelligence. Considering the difference of data observation and training principles, the Stacking based load forecasting model embedded various machine learning algorithms was proposed to utilize their diversified strength.The XGBoost algorithm constructed by tree ensemble model and long-short term memory were involved in Stacking base-learner layer. Swiss load data in ENTSO was used to verify the feasibility of the algorithm in case study. The forecasting results show that the XGBoost, the gradient decision tree and the random forest model can quantify the contribution of the input data through the gain of their own model. The load forecasting results are more accurate when each of base-learner has lower correlation coefficient in Stacking. The results indicate the Stacking ensemble learning based on multi-model has better prediction performance compared with the traditional single model.
引文
[1]薛禹胜,赖业宁.大能源思维与大数据思维的融合:(一)大数据与电力大数据[J].电力系统自动化,2016,40(1):1-8.Xue Yusheng,Lai Yening.Integration of macro energy thinking and big data thinking:part one big data and power big data[J].Automation of Electric Power Systems,2016,40(1):1-8(in Chinese).
[2]王继业,季知祥,史梦洁,等.智能配用电大数据需求分析与应用研究[J].中国电机工程学报,2015,35(8):1829-1836.Wang Jiye,Ji Zhixiang,Shi Mengjie,et al.Scenario analysis and application research on big data in smart power distribution and consumption systems[J].Proceedings of the CSEE,2015,35(8):1829-1836(in Chinese).
[3]尚宇炜,马钊,彭晨阳,等.内嵌专业知识和经验的机器学习方法探索(一):引导学习的提出与理论基础[J].中国电机工程学报,2017,37(19):5560-5571.Shang Yuwei,Ma Zhao,Peng Chenyang,et al.Study of a novel machine learning method embedding expertise part I:proposals and fundamentals of guiding learning[J].Proceedings of the CSEE,2017,37(19):5560-5571(in Chinese).
[4]史佳琪,张建华.基于深度学习的超短期光伏精细化预测模型研究[J].电力建设,2017,38(6):28-35.Shi Jiaqi,Zhang Jianhua.Ultra short-term photovoltaic refined forecasting model based on deep learning[J].Electric Power Construction,2017,38(6):28-35(in Chinese).
[5]魏东,龚庆武,来文青,等.基于卷积神经网络的输电线路区内外故障判断及故障选相方法研究[J].中国电机工程学报,2016,36(S1):21-28.Wei Dong,Gong Qingwu,Lai Wenqing,et al.Research on internal and external fault diagnosis and fault-selection of transmission line based on convolutional neural network[J].Proceedings of the CSEE,2016,36(S1):21-28(in Chinese).
[6]国务院.国务院关于印发新一代人工智能发展规划的通知[R].北京:国务院,2017.State Council.Circular of the state council on the issuance of a new generation of artificial intelligence development plan[R].Beijing:State Council,2017(in Chinese).
[7]张素香,赵丙镇,王风雨,等.海量数据下的电力负荷短期预测[J].中国电机工程学报,2015,35(1):37-42.Zhang Suxiang,Zhao Bingzhen,Wang Fengyu,et al.Short-term power load forecasting based on big data[J].Proceedings of the CSEE,2015,35(1):37-42(in Chinese).
[8]康重庆,牟涛,夏清.电力系统多级负荷预测及其协调问题:(一)研究框架[J].电力系统自动化,2008,32(7):34-38.Kang Chongqing,Mu Tao,Xia Qing.Power system multilevel load forecasting and coordinating part one:research framework[J].Automation of Electric Power Systems,2008,32(7):34-38(in Chinese).
[9]Khotanzad A,Afkhami-Rohani R,Maratukulam D.ANNSTLF-artificial neural network short-term load forecaster-generation three[J].IEEE Transactions on Power Systems,1998,13(4):1413-1422.
[10]谢敏,邓佳梁,吉祥,等.基于信息熵和变精度粗糙集优化的支持向量机降温负荷预测方法[J].电网技术,2017,41(1):210-214.Xie Min,Deng Jialiang,Ji Xiang,et al.Cooling load forecasting method based on support vector machine optimized with entropy and variable accuracy roughness set[J].Power System Technology,2017,41(1):210-214(in Chinese).
[11]沈沉,秦建,盛万兴,等.基于小波聚类的配变短期负荷预测方法研究[J].电网技术,2016,40(2):521-526.Shen Chen,Qin Jian,Sheng Wanxing,et al.Study on short-term forecasting of distribution transformer load using wavelet and clustering method[J].Power System Technology,2016,40(2):521-526(in Chinese).
[12]吴潇雨,和敬涵,张沛,等.基于灰色投影改进随机森林算法的电力系统短期负荷预测[J].电力系统自动化,2015,39(12):50-55.Wu Xiaoyu,He Jinghan,Zhang Pei,et al.Power system short-term load forecasting based on improved random forest with grey relation projection[J].Automation of Electric Power Systems,2015,39(12):50-55(in Chinese).
[13]Li Guangye,Li Wei,Tian Xiaolei,et al.Short-term electricity load forecasting based on the XGBoost algorithm[J].Smart Grid,2017,7(4):274-285.
[14]Ryu S,Noh J,Kim H.Deep neural network based demand side short term load forecasting[J].Energies,2017,10(1):3.
[15]Kong Weicong,Dong Zhaoyang,Jia Youwei,et al.Short-term residential load forecasting based on LSTMrecurrent neural network[J].IEEE Transactions on Smart Grid,2019,10(1):841-851.
[16]梁智,孙国强,李虎成,等.基于VMD与PSO优化深度信念网络的短期负荷预测[J].电网技术,2018,42(2):598-606.Liang Zhi,Sun Guoqiang,Li Hucheng,et al.Short-term load forecasting based on VMD and PSO optimized deep belief network[J].Power System Technology,2018,42(2):598-606(in Chinese).
[17]史佳琪,谭涛,郭经,等.基于深度结构多任务学习的园区型综合能源系统多元负荷预测[J].电网技术,2018,42(3):698-706.Shi Jiaqi,Tan Tao,Guo Jing,et al.Multi-task learning based on deep architecture for various types of load forecasting in regional energy system integration[J].Power System Technology,2018,42(3):698-706(in Chinese).
[18]刘文霞,龙日尚,徐晓波,等.考虑数据新鲜度和交叉熵的电动汽车短期充电负荷预测模型[J].电力系统自动化,2016,40(12):45-52.Liu Wenxia,Long Rishang,Xu Xiaobo,et al.Forecasting model of short-term EV charging load based on data freshness and cross entropy[J].Automation of Electric Power Systems,2016,40(12):45-52(in Chinese).
[19]Li Zhiyi,Liu Xuan,Chen Liyuan.Load interval forecasting methods based on an ensemble of Extreme Learning Machines[C]//Proceedings of 2015 IEEE Power&Energy Society General Meeting.Denver,CO,USA:IEEE,2015:1-5.
[20]Zhang Rui,Dong Zhaoyang,Xu Yan,et al.Short-term load forecasting of Australian national electricity market by an ensemble model of extreme learning machine[J].IET Generation,Transmission&Distribution,2013,7(4):391-397.
[21]Wolpert D H.Stacked generalization[M].Boston:Springer,2017:6-10.
[2]王继业,季知祥,史梦洁,等.智能配用电大数据需求分析与应用研究[J].中国电机工程学报,2015,35(8):1829-1836.Wang Jiye,Ji Zhixiang,Shi Mengjie,et al.Scenario analysis and application research on big data in smart power distribution and consumption systems[J].Proceedings of the CSEE,2015,35(8):1829-1836(in Chinese).
[3]尚宇炜,马钊,彭晨阳,等.内嵌专业知识和经验的机器学习方法探索(一):引导学习的提出与理论基础[J].中国电机工程学报,2017,37(19):5560-5571.Shang Yuwei,Ma Zhao,Peng Chenyang,et al.Study of a novel machine learning method embedding expertise part I:proposals and fundamentals of guiding learning[J].Proceedings of the CSEE,2017,37(19):5560-5571(in Chinese).
[4]史佳琪,张建华.基于深度学习的超短期光伏精细化预测模型研究[J].电力建设,2017,38(6):28-35.Shi Jiaqi,Zhang Jianhua.Ultra short-term photovoltaic refined forecasting model based on deep learning[J].Electric Power Construction,2017,38(6):28-35(in Chinese).
[5]魏东,龚庆武,来文青,等.基于卷积神经网络的输电线路区内外故障判断及故障选相方法研究[J].中国电机工程学报,2016,36(S1):21-28.Wei Dong,Gong Qingwu,Lai Wenqing,et al.Research on internal and external fault diagnosis and fault-selection of transmission line based on convolutional neural network[J].Proceedings of the CSEE,2016,36(S1):21-28(in Chinese).
[6]国务院.国务院关于印发新一代人工智能发展规划的通知[R].北京:国务院,2017.State Council.Circular of the state council on the issuance of a new generation of artificial intelligence development plan[R].Beijing:State Council,2017(in Chinese).
[7]张素香,赵丙镇,王风雨,等.海量数据下的电力负荷短期预测[J].中国电机工程学报,2015,35(1):37-42.Zhang Suxiang,Zhao Bingzhen,Wang Fengyu,et al.Short-term power load forecasting based on big data[J].Proceedings of the CSEE,2015,35(1):37-42(in Chinese).
[8]康重庆,牟涛,夏清.电力系统多级负荷预测及其协调问题:(一)研究框架[J].电力系统自动化,2008,32(7):34-38.Kang Chongqing,Mu Tao,Xia Qing.Power system multilevel load forecasting and coordinating part one:research framework[J].Automation of Electric Power Systems,2008,32(7):34-38(in Chinese).
[9]Khotanzad A,Afkhami-Rohani R,Maratukulam D.ANNSTLF-artificial neural network short-term load forecaster-generation three[J].IEEE Transactions on Power Systems,1998,13(4):1413-1422.
[10]谢敏,邓佳梁,吉祥,等.基于信息熵和变精度粗糙集优化的支持向量机降温负荷预测方法[J].电网技术,2017,41(1):210-214.Xie Min,Deng Jialiang,Ji Xiang,et al.Cooling load forecasting method based on support vector machine optimized with entropy and variable accuracy roughness set[J].Power System Technology,2017,41(1):210-214(in Chinese).
[11]沈沉,秦建,盛万兴,等.基于小波聚类的配变短期负荷预测方法研究[J].电网技术,2016,40(2):521-526.Shen Chen,Qin Jian,Sheng Wanxing,et al.Study on short-term forecasting of distribution transformer load using wavelet and clustering method[J].Power System Technology,2016,40(2):521-526(in Chinese).
[12]吴潇雨,和敬涵,张沛,等.基于灰色投影改进随机森林算法的电力系统短期负荷预测[J].电力系统自动化,2015,39(12):50-55.Wu Xiaoyu,He Jinghan,Zhang Pei,et al.Power system short-term load forecasting based on improved random forest with grey relation projection[J].Automation of Electric Power Systems,2015,39(12):50-55(in Chinese).
[13]Li Guangye,Li Wei,Tian Xiaolei,et al.Short-term electricity load forecasting based on the XGBoost algorithm[J].Smart Grid,2017,7(4):274-285.
[14]Ryu S,Noh J,Kim H.Deep neural network based demand side short term load forecasting[J].Energies,2017,10(1):3.
[15]Kong Weicong,Dong Zhaoyang,Jia Youwei,et al.Short-term residential load forecasting based on LSTMrecurrent neural network[J].IEEE Transactions on Smart Grid,2019,10(1):841-851.
[16]梁智,孙国强,李虎成,等.基于VMD与PSO优化深度信念网络的短期负荷预测[J].电网技术,2018,42(2):598-606.Liang Zhi,Sun Guoqiang,Li Hucheng,et al.Short-term load forecasting based on VMD and PSO optimized deep belief network[J].Power System Technology,2018,42(2):598-606(in Chinese).
[17]史佳琪,谭涛,郭经,等.基于深度结构多任务学习的园区型综合能源系统多元负荷预测[J].电网技术,2018,42(3):698-706.Shi Jiaqi,Tan Tao,Guo Jing,et al.Multi-task learning based on deep architecture for various types of load forecasting in regional energy system integration[J].Power System Technology,2018,42(3):698-706(in Chinese).
[18]刘文霞,龙日尚,徐晓波,等.考虑数据新鲜度和交叉熵的电动汽车短期充电负荷预测模型[J].电力系统自动化,2016,40(12):45-52.Liu Wenxia,Long Rishang,Xu Xiaobo,et al.Forecasting model of short-term EV charging load based on data freshness and cross entropy[J].Automation of Electric Power Systems,2016,40(12):45-52(in Chinese).
[19]Li Zhiyi,Liu Xuan,Chen Liyuan.Load interval forecasting methods based on an ensemble of Extreme Learning Machines[C]//Proceedings of 2015 IEEE Power&Energy Society General Meeting.Denver,CO,USA:IEEE,2015:1-5.
[20]Zhang Rui,Dong Zhaoyang,Xu Yan,et al.Short-term load forecasting of Australian national electricity market by an ensemble model of extreme learning machine[J].IET Generation,Transmission&Distribution,2013,7(4):391-397.
[21]Wolpert D H.Stacked generalization[M].Boston:Springer,2017:6-10.