基于FA-PCA-LSTM的光伏发电短期功率预测
|
摘要
准确的短期光伏功率预测是调度部门合理制定发电计划、保证电力系统安全性和经济性的关键性技术.针对光伏出力可预测性低的问题,提出了一种结合因子分析(factor analysis,FA)、主成分分析(principal component analysis,PCA)和长短期记忆网络(long short-term memory,LSTM)的光伏发电短期功率预测方法.首先采用因子分析对多元数据序列信息进行分析,提取相关性较高的公共因子并优化样本.然后通过主成分分析对优化后的多元数据序列进行筛选,在充分利用序列信息的基础上降低数据规模和复杂程度.最后,利用LSTM网络对多元数据序列与光伏功率序列之间的非线性关系进行动态时间建模并预测.采用中国新疆某光伏电站的实测数据进行验证,算例分析结果表明所提预测方法的有效性.
Accurate short-term photovoltaic power forecasting is the key technology for dispatching department to make reasonable generation plan and ensure the security and economy of power system. Considering poor predictability of PV output,this paper propose a short-term power forecasting method for photovoltaic power generation based on factor analysis( FA),principal component analysis( PCA) and long short-term memory( LSTM).Firstly,factor analysis is used to analyze the multivariate data sequence information,and common factors with high correlation are extracted and samples are optimized. Then the optimized multivariate data sequence is screened by principal component analysis,which reduces the size and complexity of data on the basis of making full use of sequence information. Finally,LSTM network is used to model and predict the nonlinear relationship between multivariate data series and photovoltaic power series. Verified by the measured data of one photovoltaic power plant in Xinjiang,China,the calculation results show the effectiveness of the proposed prediction method.
Accurate short-term photovoltaic power forecasting is the key technology for dispatching department to make reasonable generation plan and ensure the security and economy of power system. Considering poor predictability of PV output,this paper propose a short-term power forecasting method for photovoltaic power generation based on factor analysis( FA),principal component analysis( PCA) and long short-term memory( LSTM).Firstly,factor analysis is used to analyze the multivariate data sequence information,and common factors with high correlation are extracted and samples are optimized. Then the optimized multivariate data sequence is screened by principal component analysis,which reduces the size and complexity of data on the basis of making full use of sequence information. Finally,LSTM network is used to model and predict the nonlinear relationship between multivariate data series and photovoltaic power series. Verified by the measured data of one photovoltaic power plant in Xinjiang,China,the calculation results show the effectiveness of the proposed prediction method.
引文
[1]北极星太阳能光伏网.2018年中国光伏产业发展形势展望[EB/OL].http://guangfu.bjx.com.cn/news/20171228/870807.shtml.
[2]Zhou Z,Xiong F,Huang B,et al.Game-theoretical energy management for energy internet with big data-based renewable power forecasting[J].IEEE Access,2017,5(99):5731-5746.
[3]Li Y,Su Y,Shu L.An ARMAX model for forecasting the power output of a grid connected photovoltaic system[J].Renewable Energy,2014,66(6):78-89.
[4]Voyant C,Muselli M,Paoli C,et al.Numerical weather prediction(NWP)and hybrid ARMA/ANN model to predict global radiation[J].Energy,2012,39(1):341-355.
[5]黄磊,舒杰,姜桂秀,等.基于多维时间序列局部支持向量回归的微网光伏发电预测[J].电力系统自动化,2014,38(5):19-24.
[6]李国庆,张钰,张明江,等.基于MRMR的集合经验模态分解和支持向量机的风电功率实时预测[J]东北电力大学学报,2017,37(2):39-44.
[7]袁晓玲,施俊华,徐杰彦.计及天气类型指数的光伏发电短期出力预测[J].中国电机工程学报,2013,33(34):57-64.
[8]杨茂,黄宾阳,江博,等.基于卡尔曼滤波和支持向量机的风电功率实时预测研究[J].东北电力大学学报,2017,37(2):45-51.
[9]董雷,周文萍,张沛,等.基于动态贝叶斯网络的光伏发电短期概率预测[J].中国电机工程学报,2013,33(S1):38-45.
[10]叶林,陈政,赵永宁,等.基于遗传算法-模糊径向基神经网络的光伏发电功率预测模型[J].电力系统自动化,2015,39(16):16-22.
[11]赵唯嘉,张宁,康重庆,等.光伏发电出力的条件预测误差概率分布估计方法[J].电力系统自动化,2015,39(16):8-15.
[12]Silva T V D,Monteiro R V A,Moura F A M,et al.Performance analysis of neural network training algorithms and support vector machine for power generation forecast of photovoltaic panel[J].IEEE Latin America Transactions,2017,15(6):1091-1100.
[13]陈志宝,丁杰,周海,等.地基云图结合径向基函数人工神经网络的光伏功率超短期预测模型[J].中国电机工程学报,2015,35(3):561-567.
[14]龚莺飞,鲁宗相,乔颖,等.光伏功率预测技术[J].电力系统自动化,2016,40(4):140-151.
[15]Liu J,Fang W,Zhang X,et al.An improved photovoltaic power forecasting model with the assistance of aerosol index data[J].IEEE Transactions on Sustainable Energy,2015,6(2):434-442.
[16]Shi J,Lee W J,Liu Y,et al.Forecasting power output of photovoltaic systems based on weather classification and support vector machines[J].IEEE Transactions on Industry Applications,2015,48(3):1064-1069.
[17]Yang H T,Huang C M,Huang Y C,et al.A weather-based hybrid method for 1-day ahead hourly forecasting of PV power output[J].IEEE Transactions on Sustainable Energy,2014,5(3):917-926.
[18]邱革非,张春刚,仲泽坤,等.光伏发电并网系统相量模型建模研究[J].昆明理工大学学报(自然科学版),2015,40(4):57-62.
[19]刘卫亮,刘长良,林永君,等.计及雾霾影响因素的光伏发电超短期功率预测[J].中国电机工程学报,2018,38(14):4086-4095.
[20]Chen K,Huo Q.Training deep bidirectional LSTM acoustic model for LVCSR by a context-sensitive-chunk BPTT approach[J].IEEE/ACM Transactions on Audio,Speech and Language Processing,2016,24(7):1185-1193.
[21]Song E,Song F K,Kang H G.Effective spectral and excitation modeling techniques for LSTM-RNN-Based speech synthesis systems[J].IEEE/ACM Transactions on Audio,Speech and Language Processing,2017,25(11):2152-2161.
[22]Gao L,Guo Z,Zhang H,et al.Video captioning with attention-based LSTM and semantic consistency[J].IEEE Transactions on Multimedia,2017,19(9):2045-2055.
[23]朱乔木,李弘毅,王子琪,等.基于长短期记忆网络的风电场发电功率超短期预测[J].电网技术,2017,41(12):3797-3802.
[24]郁琛,薛禹胜,文福拴,等.风电功率预测误差的风险评估[J].电力系统自动化,2015,39(7):52-58.
[2]Zhou Z,Xiong F,Huang B,et al.Game-theoretical energy management for energy internet with big data-based renewable power forecasting[J].IEEE Access,2017,5(99):5731-5746.
[3]Li Y,Su Y,Shu L.An ARMAX model for forecasting the power output of a grid connected photovoltaic system[J].Renewable Energy,2014,66(6):78-89.
[4]Voyant C,Muselli M,Paoli C,et al.Numerical weather prediction(NWP)and hybrid ARMA/ANN model to predict global radiation[J].Energy,2012,39(1):341-355.
[5]黄磊,舒杰,姜桂秀,等.基于多维时间序列局部支持向量回归的微网光伏发电预测[J].电力系统自动化,2014,38(5):19-24.
[6]李国庆,张钰,张明江,等.基于MRMR的集合经验模态分解和支持向量机的风电功率实时预测[J]东北电力大学学报,2017,37(2):39-44.
[7]袁晓玲,施俊华,徐杰彦.计及天气类型指数的光伏发电短期出力预测[J].中国电机工程学报,2013,33(34):57-64.
[8]杨茂,黄宾阳,江博,等.基于卡尔曼滤波和支持向量机的风电功率实时预测研究[J].东北电力大学学报,2017,37(2):45-51.
[9]董雷,周文萍,张沛,等.基于动态贝叶斯网络的光伏发电短期概率预测[J].中国电机工程学报,2013,33(S1):38-45.
[10]叶林,陈政,赵永宁,等.基于遗传算法-模糊径向基神经网络的光伏发电功率预测模型[J].电力系统自动化,2015,39(16):16-22.
[11]赵唯嘉,张宁,康重庆,等.光伏发电出力的条件预测误差概率分布估计方法[J].电力系统自动化,2015,39(16):8-15.
[12]Silva T V D,Monteiro R V A,Moura F A M,et al.Performance analysis of neural network training algorithms and support vector machine for power generation forecast of photovoltaic panel[J].IEEE Latin America Transactions,2017,15(6):1091-1100.
[13]陈志宝,丁杰,周海,等.地基云图结合径向基函数人工神经网络的光伏功率超短期预测模型[J].中国电机工程学报,2015,35(3):561-567.
[14]龚莺飞,鲁宗相,乔颖,等.光伏功率预测技术[J].电力系统自动化,2016,40(4):140-151.
[15]Liu J,Fang W,Zhang X,et al.An improved photovoltaic power forecasting model with the assistance of aerosol index data[J].IEEE Transactions on Sustainable Energy,2015,6(2):434-442.
[16]Shi J,Lee W J,Liu Y,et al.Forecasting power output of photovoltaic systems based on weather classification and support vector machines[J].IEEE Transactions on Industry Applications,2015,48(3):1064-1069.
[17]Yang H T,Huang C M,Huang Y C,et al.A weather-based hybrid method for 1-day ahead hourly forecasting of PV power output[J].IEEE Transactions on Sustainable Energy,2014,5(3):917-926.
[18]邱革非,张春刚,仲泽坤,等.光伏发电并网系统相量模型建模研究[J].昆明理工大学学报(自然科学版),2015,40(4):57-62.
[19]刘卫亮,刘长良,林永君,等.计及雾霾影响因素的光伏发电超短期功率预测[J].中国电机工程学报,2018,38(14):4086-4095.
[20]Chen K,Huo Q.Training deep bidirectional LSTM acoustic model for LVCSR by a context-sensitive-chunk BPTT approach[J].IEEE/ACM Transactions on Audio,Speech and Language Processing,2016,24(7):1185-1193.
[21]Song E,Song F K,Kang H G.Effective spectral and excitation modeling techniques for LSTM-RNN-Based speech synthesis systems[J].IEEE/ACM Transactions on Audio,Speech and Language Processing,2017,25(11):2152-2161.
[22]Gao L,Guo Z,Zhang H,et al.Video captioning with attention-based LSTM and semantic consistency[J].IEEE Transactions on Multimedia,2017,19(9):2045-2055.
[23]朱乔木,李弘毅,王子琪,等.基于长短期记忆网络的风电场发电功率超短期预测[J].电网技术,2017,41(12):3797-3802.
[24]郁琛,薛禹胜,文福拴,等.风电功率预测误差的风险评估[J].电力系统自动化,2015,39(7):52-58.