基于自适应深度信念网络的变电站负荷预测
|
摘要
精确的变电站级负荷预测是电网精益化运行决策的重要基础,但存在不同站间负荷特性差异大、微观关联因素多样性强等传统预测方法难以处理的问题。基于变电站历史负荷数据与其所在区域的外部环境数据,通过深度信念网络算法(deep belief network,DBN)强大的学习能力,避免了相似日等特征选取问题,并采用Nadam动量优化算法训练深度信念网络,得到DBN最佳参数,构成针对变电站负荷预测的学习框架,并基于Keras深度学习框架自动调整DBN结构,达到最优预测结果。以20个具有典型负荷特性的220kV变电站实际负荷数据为样本集,在周、日和小时级3个预测时间尺度上,通过2种误差计算方式作实例对比证明,所提方法能够充分进行自适应深度学习,并进行高精度变电站级负荷预测。
Precise substation-level load forecasting is an important basis for grid lean operation decision-making, but there are problems such as large differences in load characteristics between different substations, and diversified micro-correlation factors that are difficult to deal with by traditional forecasting methods. Based on the historical load data of the substation and the external environmental data of the area where it is located, this paper avoided the selection of features such as similar days through the strong learning ability of deep belief network, and adopted the Nadam momentum optimization algorithm to train the deep belief network and obtain the best DBN parameters to construct a learning framework for substation load forecasting, and automatically adjusted the parameters based on the framework of Keras deep learning framework to achieve the optimal forecast results.Take the actual operation load data of 20 220 kV substations with typical load characteristics as a sample set, and compare it with two error calculation methods on the three prediction time scales of week, day and hour to prove that the proposed method can fully implement adaptive depth learning and perform high-precision substation-level load forecasting.
Precise substation-level load forecasting is an important basis for grid lean operation decision-making, but there are problems such as large differences in load characteristics between different substations, and diversified micro-correlation factors that are difficult to deal with by traditional forecasting methods. Based on the historical load data of the substation and the external environmental data of the area where it is located, this paper avoided the selection of features such as similar days through the strong learning ability of deep belief network, and adopted the Nadam momentum optimization algorithm to train the deep belief network and obtain the best DBN parameters to construct a learning framework for substation load forecasting, and automatically adjusted the parameters based on the framework of Keras deep learning framework to achieve the optimal forecast results.Take the actual operation load data of 20 220 kV substations with typical load characteristics as a sample set, and compare it with two error calculation methods on the three prediction time scales of week, day and hour to prove that the proposed method can fully implement adaptive depth learning and perform high-precision substation-level load forecasting.
引文
[1]刘国静,韩学山,杨明.电力系统运行协同的经济调度[J].中国电机工程学报,2014,34(16):2668-2675.Liu Guojing,Han Xueshan,Yang Ming.Synergetic economic dispatch in power system operation[J].Proceedings of the CSEE,2014,34(16):2668-2675(in Chinese).
[2]雷雪姣,潘士娟,管晓宏,等.考虑传输安全裕度的电力系统发电经济调度[J].中国电机工程学报,2014,34(31):5651-5658.Lei Xuejiao,Pan Shijuan,Guan Xiaohong,et al.Transmission safety margin constrained unit commitment in power systems[J].Proceedings of the CSEE,2014,34(31):5651-5658(in Chinese).
[3]廖旎焕,胡智宏,马莹莹,等.电力系统短期负荷预测方法综述[J].电力系统保护与控制,2011,39(1):147-152.Liao Nihuan,Hu Zhihong,Ma Yingying,et al.Review of the short-term load forecasting methods of electric power system[J].Power System Protection and Control,2011,39(1):147-152(in Chinese).
[4]周孝信,陈树勇,鲁宗相,等.能源转型中我国新一代电力系统的技术特征[J].中国电机工程学报,2018,38(7):1893-1904.Zhou Xiaoxin,Chen Shuyong,Lu Zongxiang,et al.Technology features of the new generation power system in china[J].Proceedings of the CSEE,2018,38(7):1893-1904(in Chinese).
[5]鲁宗相,李海波,乔颖.高比例可再生能源并网的电力系统灵活性评价与平衡机理[J].中国电机工程学报,2017,37(1):9-19.Lu Zongxiang,Li Haibo,Qiao Ying.Flexibility evaluation and supply/demand balance principle of power system with high-penetration renewable electricity[J].Proceedings of the CSEE,2017,37(1):9-19(in Chinese).
[6]高红均,刘俊勇.考虑不同类型DG和负荷建模的主动配电网协同规划[J].中国电机工程学报,2016,36(18):4911-4922.Gao Hongjun,Liu Junyong.Coordinated planning considering different types of DG and load in active distribution network[J].Proceedings of the CSEE,2016,36(18):4911-4922(in Chinese).
[7]杨威,向月,刘俊勇,等.基于多代理技术的电动汽车规模演化模型[J].电网技术,2017,41(7):2146-2154.Yang Wei,Xiang Yue,Liu Junyong,et al.Multi-agent modeling for the scale evolution of plug-in electric vehicles[J].Power System Technology,2017,41(7):2146-2154(in Chinese).
[8]胡泽春,宋永华,徐智威,等.电动汽车接入电网的影响与利用[J].中国电机工程学报,2012,32(4):1-10.Hu Zechun,Song Yonghua,Xu Zhiwei,et al.Impacts and utilization of electric vehicles integration into power systems[J].Proceedings of the CSEE,2012,32(4):1-10(in Chinese).
[9]Chen J F,Wang Weiming,Huang Chaoming.Analysis of an adaptive time-series autoregressive moving-average(ARMA)model for short-term load forecasting[J].Electric Power Systems Research,1995,34(3):187-196.
[10]Hagan M T,Behr S M.The time series approach to short term load forecasting[J].IEEE Transactions on Power Systems,1987,2(3):785-791.
[11]Huang Li,Yang Yongbiao,Zhao Honglei,et al.Time series modeling and filtering method of electric power load stochastic noise[J].Protection and Control of Modern Power Systems,2017(2):25.
[12]谷云东,张素杰,冯君淑.大用户电力负荷的多模型模糊综合预测[J].电工技术学报,2015,30(23):110-115.Gu Yundong,Zhang Sujie,Feng Junshu.Multi-model fuzzy synthesis forecasting of electric power loads for larger consumers[J].Transactions of China Electrotechnical Society,2015,30(23):110-115(in Chinese).
[13]杜鹃.改进的TSK型动态模糊神经网络在短期负荷预测中的应用[J].电网技术,2010,34(4):69-75.Du Juan.An improved TSK-type dynamic fuzzy neural network approach for short-term load forecasting[J].Power System Technology,2010,34(4):69-75(in Chinese).
[14]熊军华,牛珂,张春歌,等.基于小波变异果蝇优化支持向量机短期负荷预测方法研究[J].电力系统保护与控制,2017,45(13):71-77.Xiong Junhua,Niu Ke,Zhang Chunge,et al.LSSVM in short-term load forecasting based on wavelet transform and mutant fruit fly optimization algorithm[J].Power System Protection and Control,2017,45(13):71-77(in Chinese).
[15]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).
[16]肖白,聂鹏,穆钢,等.基于多级聚类分析和支持向量机的空间负荷预测方法[J].电力系统自动化,2015,39(12):56-61.Xiao Bai,Nie Peng,Mu Gang,et al.A spatial load forecasting method based on multilevel clustering analysis and support vector machine[J].Automation of Electric Power Systems,2015,39(12):56-61(in Chinese).
[17]Huang Nantian,Lu Guobo,Xu Dianguo.A Permutation Importance-based feature selection method for short-term electricity load forecasting using random forest[J].Energies,2016,9(10):767.
[18]Lahouar A,Slama J B H.Day-ahead load forecast using random forest and expert input selection[J].Energy Conversion and Management,2015,103:1040-1051.
[19]陈亮,王震,王刚.深度学习框架下LSTM网络在短期电力负荷预测中的应用[J].电力信息与通信技术,2017,15(5):8-11.Chen Liang,Wang Zhen,Wang Gang.Application of LSTM networks in short-term power load forecasting under the deep learning framework[J].Electric Power Information and Communication Technology,2017,15(5):8-11(in Chinese).
[20]张宇航,邱才明,贺兴,等.一种基于LSTM神经网络的短期用电负荷预测方法[J].电力信息与通信技术,2017,15(9):19-25.Zhang Yuhang,Qiu Caiming,He Xing,et al.A short-term load forecasting based on LSTM neural network[J].Electric Power Information and Communication Technology,2017,15(9):19-25(in Chinese).
[21]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).
[22]牛东晓,谷志红,邢棉,等.基于数据挖掘的SVM短期负荷预测方法研究[J].中国电机工程学报,2006,26(18):6-12.Niu Dongxiao,Gu Zhihong,Xing Mian,et al.Study on forecasting approach to short-term load of SVM based on data mining[J].Proceedings of the CSEE,2006,26(18):6-12(in Chinese).
[23]孔祥玉,郑锋,鄂志君,等.基于深度信念网络的短期负荷预测方法[J].电力系统自动化,2018,42(5):133-139.Kong Xiangyu,Zheng Feng,E Zhijun,et al.Short-term load forecasting based on deep belief network[J].Automation of Electric Power Systems,2018,42(5):133-139(in Chinese).
[24]李龙,魏靖,黎灿兵,等.基于人工神经网络的负荷模型预测[J].电工技术学报,2015,30(8):225-230.Li Long,Wei Jing,Li Canbing,et al.Prediction of load model based on artificial neural network[J].Transactions of China Electrotechnical Society,2015,30(8):225-230(in Chinese).
[25]Mordjaoui M,Haddad S,Medoued A,et al.Electric load forecasting by using dynamic neural network[J].International Journal of Hydrogen Energy,2017,42(28):17655-17663.
[26]张智晟,于道林.考虑需求响应综合影响因素的RBF-NN短期负荷预测模型[J].中国电机工程学报,2018,38(6):1631-1638.Zhang Zhisheng,Yu Daolin.RBF-NN based short-term load forecasting model considering comprehensive factors affecting demand response[J].Proceedings of the CSEE,2018,38(6):1631-1638(in Chinese).
[27]Kong Weicong,Dong Zhaoyang,Jia Youwei.Short-term residential load forecasting based on LSTM recurrent neural network[J].IEEE Transactions on Smart Grid,2017,10(1):841-851.
[28]王琦,张文朝,汤勇,等.统计综合法负荷建模中的调查方法及应用[J].电网技术,2010,34(2):104-108.Wang Qi,Zhang Wenchao,Tang Yong,et al.A new load survey method and its application in component based load modeling[J].Power System Technology,2010,34(2):104-108(in Chinese).
[29]Hinton G E,Osindero S,Teh Y W.A fast learning algorithm for deep belief nets[J].Neural Computation,2006,18(7):1527-1554.
[30]Hinton G E.Training products of experts by minimizing contrastive divergence[J].Neural Computation,2002,14(8):1771-1800.
[31]Hinton G.A practical guide to training restricted Boltzmann machines[J].Momentum,2010,9(1):926-947.
[32]Diederik P.Kingma,Jimmy Ba.Adam:A Method for Stochastic Optimization[C]//Proceedings of International Conference on Learning Representations.San Diego,CA,America,2015.
[33]McMahan B,Streeter M.Delay-tolerant algorithms for asynchronous distributed online learning[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems.Montreal,Canada:MITPress Cambridge,2014:2915-2923.
[34]Sutskever I,Martens J,Dahl G,et al.On the importance of initialization and momentum in deep learning[C]//Proceedings of the 30th International Conference on International Conference on Machine Learning.Atlanta,GA,USA:JMLR,2013,38(5):III-1139-III-1147.
[2]雷雪姣,潘士娟,管晓宏,等.考虑传输安全裕度的电力系统发电经济调度[J].中国电机工程学报,2014,34(31):5651-5658.Lei Xuejiao,Pan Shijuan,Guan Xiaohong,et al.Transmission safety margin constrained unit commitment in power systems[J].Proceedings of the CSEE,2014,34(31):5651-5658(in Chinese).
[3]廖旎焕,胡智宏,马莹莹,等.电力系统短期负荷预测方法综述[J].电力系统保护与控制,2011,39(1):147-152.Liao Nihuan,Hu Zhihong,Ma Yingying,et al.Review of the short-term load forecasting methods of electric power system[J].Power System Protection and Control,2011,39(1):147-152(in Chinese).
[4]周孝信,陈树勇,鲁宗相,等.能源转型中我国新一代电力系统的技术特征[J].中国电机工程学报,2018,38(7):1893-1904.Zhou Xiaoxin,Chen Shuyong,Lu Zongxiang,et al.Technology features of the new generation power system in china[J].Proceedings of the CSEE,2018,38(7):1893-1904(in Chinese).
[5]鲁宗相,李海波,乔颖.高比例可再生能源并网的电力系统灵活性评价与平衡机理[J].中国电机工程学报,2017,37(1):9-19.Lu Zongxiang,Li Haibo,Qiao Ying.Flexibility evaluation and supply/demand balance principle of power system with high-penetration renewable electricity[J].Proceedings of the CSEE,2017,37(1):9-19(in Chinese).
[6]高红均,刘俊勇.考虑不同类型DG和负荷建模的主动配电网协同规划[J].中国电机工程学报,2016,36(18):4911-4922.Gao Hongjun,Liu Junyong.Coordinated planning considering different types of DG and load in active distribution network[J].Proceedings of the CSEE,2016,36(18):4911-4922(in Chinese).
[7]杨威,向月,刘俊勇,等.基于多代理技术的电动汽车规模演化模型[J].电网技术,2017,41(7):2146-2154.Yang Wei,Xiang Yue,Liu Junyong,et al.Multi-agent modeling for the scale evolution of plug-in electric vehicles[J].Power System Technology,2017,41(7):2146-2154(in Chinese).
[8]胡泽春,宋永华,徐智威,等.电动汽车接入电网的影响与利用[J].中国电机工程学报,2012,32(4):1-10.Hu Zechun,Song Yonghua,Xu Zhiwei,et al.Impacts and utilization of electric vehicles integration into power systems[J].Proceedings of the CSEE,2012,32(4):1-10(in Chinese).
[9]Chen J F,Wang Weiming,Huang Chaoming.Analysis of an adaptive time-series autoregressive moving-average(ARMA)model for short-term load forecasting[J].Electric Power Systems Research,1995,34(3):187-196.
[10]Hagan M T,Behr S M.The time series approach to short term load forecasting[J].IEEE Transactions on Power Systems,1987,2(3):785-791.
[11]Huang Li,Yang Yongbiao,Zhao Honglei,et al.Time series modeling and filtering method of electric power load stochastic noise[J].Protection and Control of Modern Power Systems,2017(2):25.
[12]谷云东,张素杰,冯君淑.大用户电力负荷的多模型模糊综合预测[J].电工技术学报,2015,30(23):110-115.Gu Yundong,Zhang Sujie,Feng Junshu.Multi-model fuzzy synthesis forecasting of electric power loads for larger consumers[J].Transactions of China Electrotechnical Society,2015,30(23):110-115(in Chinese).
[13]杜鹃.改进的TSK型动态模糊神经网络在短期负荷预测中的应用[J].电网技术,2010,34(4):69-75.Du Juan.An improved TSK-type dynamic fuzzy neural network approach for short-term load forecasting[J].Power System Technology,2010,34(4):69-75(in Chinese).
[14]熊军华,牛珂,张春歌,等.基于小波变异果蝇优化支持向量机短期负荷预测方法研究[J].电力系统保护与控制,2017,45(13):71-77.Xiong Junhua,Niu Ke,Zhang Chunge,et al.LSSVM in short-term load forecasting based on wavelet transform and mutant fruit fly optimization algorithm[J].Power System Protection and Control,2017,45(13):71-77(in Chinese).
[15]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).
[16]肖白,聂鹏,穆钢,等.基于多级聚类分析和支持向量机的空间负荷预测方法[J].电力系统自动化,2015,39(12):56-61.Xiao Bai,Nie Peng,Mu Gang,et al.A spatial load forecasting method based on multilevel clustering analysis and support vector machine[J].Automation of Electric Power Systems,2015,39(12):56-61(in Chinese).
[17]Huang Nantian,Lu Guobo,Xu Dianguo.A Permutation Importance-based feature selection method for short-term electricity load forecasting using random forest[J].Energies,2016,9(10):767.
[18]Lahouar A,Slama J B H.Day-ahead load forecast using random forest and expert input selection[J].Energy Conversion and Management,2015,103:1040-1051.
[19]陈亮,王震,王刚.深度学习框架下LSTM网络在短期电力负荷预测中的应用[J].电力信息与通信技术,2017,15(5):8-11.Chen Liang,Wang Zhen,Wang Gang.Application of LSTM networks in short-term power load forecasting under the deep learning framework[J].Electric Power Information and Communication Technology,2017,15(5):8-11(in Chinese).
[20]张宇航,邱才明,贺兴,等.一种基于LSTM神经网络的短期用电负荷预测方法[J].电力信息与通信技术,2017,15(9):19-25.Zhang Yuhang,Qiu Caiming,He Xing,et al.A short-term load forecasting based on LSTM neural network[J].Electric Power Information and Communication Technology,2017,15(9):19-25(in Chinese).
[21]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).
[22]牛东晓,谷志红,邢棉,等.基于数据挖掘的SVM短期负荷预测方法研究[J].中国电机工程学报,2006,26(18):6-12.Niu Dongxiao,Gu Zhihong,Xing Mian,et al.Study on forecasting approach to short-term load of SVM based on data mining[J].Proceedings of the CSEE,2006,26(18):6-12(in Chinese).
[23]孔祥玉,郑锋,鄂志君,等.基于深度信念网络的短期负荷预测方法[J].电力系统自动化,2018,42(5):133-139.Kong Xiangyu,Zheng Feng,E Zhijun,et al.Short-term load forecasting based on deep belief network[J].Automation of Electric Power Systems,2018,42(5):133-139(in Chinese).
[24]李龙,魏靖,黎灿兵,等.基于人工神经网络的负荷模型预测[J].电工技术学报,2015,30(8):225-230.Li Long,Wei Jing,Li Canbing,et al.Prediction of load model based on artificial neural network[J].Transactions of China Electrotechnical Society,2015,30(8):225-230(in Chinese).
[25]Mordjaoui M,Haddad S,Medoued A,et al.Electric load forecasting by using dynamic neural network[J].International Journal of Hydrogen Energy,2017,42(28):17655-17663.
[26]张智晟,于道林.考虑需求响应综合影响因素的RBF-NN短期负荷预测模型[J].中国电机工程学报,2018,38(6):1631-1638.Zhang Zhisheng,Yu Daolin.RBF-NN based short-term load forecasting model considering comprehensive factors affecting demand response[J].Proceedings of the CSEE,2018,38(6):1631-1638(in Chinese).
[27]Kong Weicong,Dong Zhaoyang,Jia Youwei.Short-term residential load forecasting based on LSTM recurrent neural network[J].IEEE Transactions on Smart Grid,2017,10(1):841-851.
[28]王琦,张文朝,汤勇,等.统计综合法负荷建模中的调查方法及应用[J].电网技术,2010,34(2):104-108.Wang Qi,Zhang Wenchao,Tang Yong,et al.A new load survey method and its application in component based load modeling[J].Power System Technology,2010,34(2):104-108(in Chinese).
[29]Hinton G E,Osindero S,Teh Y W.A fast learning algorithm for deep belief nets[J].Neural Computation,2006,18(7):1527-1554.
[30]Hinton G E.Training products of experts by minimizing contrastive divergence[J].Neural Computation,2002,14(8):1771-1800.
[31]Hinton G.A practical guide to training restricted Boltzmann machines[J].Momentum,2010,9(1):926-947.
[32]Diederik P.Kingma,Jimmy Ba.Adam:A Method for Stochastic Optimization[C]//Proceedings of International Conference on Learning Representations.San Diego,CA,America,2015.
[33]McMahan B,Streeter M.Delay-tolerant algorithms for asynchronous distributed online learning[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems.Montreal,Canada:MITPress Cambridge,2014:2915-2923.
[34]Sutskever I,Martens J,Dahl G,et al.On the importance of initialization and momentum in deep learning[C]//Proceedings of the 30th International Conference on International Conference on Machine Learning.Atlanta,GA,USA:JMLR,2013,38(5):III-1139-III-1147.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |