陈海宏,易永利,黄珅,韩静怡
CHEN Haihong,YI Yongli,HUANG Shen,HAN Jingyi

• 1:国网浙江省电力有限公司温州供电公司
• 2:亿可能源科技(上海)有限公司

摘要(Abstract)：

光伏电站发电功率的间歇性与波动性对电网安全、稳定、经济运行的影响日益明显，因此需要不断提高光伏发电功率预测准确率，为电网灵活调度与规划提供准确信息。首先，介绍了短期光伏发电功率的预测算法、特征方程、预测流程以及评价指标。接着，通过SHAP方法对训练集所构造特征进行分析筛选，使用CatBoost算法进行训练。最后，通过与使用相同特征的其他机器学习算法模型预测精度的对比，表明所提方法有效提高了预测性能，证实了基于CatBoost算法、融合多维特征的模型在光伏功率预测中的优势。
The intermittent and fluctuating generation power of PV power plants has an increasingly prominent impact on the safe, stable, and economical operation of power grids. Therefore, it is required to continuously improve the accuracy of PV power prediction to provide accurate information for flexible grid dispatching and planning.First, the prediction algorithm, characteristic equation, prediction process, and evaluation index of short-term PV generation power are introduced. Afterward, the features constructed in the training set are analyzed and filtered using the SHAP, and the training is performed using the CatBoost. Finally, by comparing the prediction accuracy with other machine learning algorithm models using the same features, the paper indicates that the proposed method can improve the prediction performance and confirms the advantages of the CatBoost that incorporates multidimensional feature models in PV power prediction.

关键词(KeyWords)： 光伏发电;功率预测;CatBoost;SHAP
PV power generation;power prediction;CatBoost;SHAP

Abstract:

Keywords:

基金项目(Foundation): 国网浙江省电力有限公司科技项目（B311WZ220002）

作者(Author): 陈海宏,易永利,黄珅,韩静怡
CHEN Haihong,YI Yongli,HUANG Shen,HAN Jingyi

DOI: 10.19585/j.zjdl.202302009

参考文献(References)：

• [1]李美成，高中亮，王龙泽，等.“双碳”目标下我国太阳能利用技术的发展现状与展望[J].太阳能，2021(11):13-18.LI Meicheng,GAO Zhongliang,WANG Longze,et al.Development status and prospect of solar energy utilization technology in China under goal of emission peak and carbon neutrality[J].Solar Energy,2021(11):13-18.
• [2]赖昌伟，黎静华，陈博，等.光伏发电出力预测技术研究综述[J].电工技术学报，2019,34(6):1201-1217.LAI Changwei,LI Jinghua,CHEN Bo,et al. Review of photovoltaic power output prediction technology[J].Transactions of China Electrotechnical Society,2019,34(6):1201-1217.
• [3]国家能源局.“十四五”现代能源体系规划[R/OL].(2022-01-29)[2022-05-24]. http：//zfxxgk. nea. gov. cn/1310524241_16479412513081n.pdf.
• [4]钱振，蔡世波，顾宇庆，等.光伏发电功率预测方法研究综述[J].机电工程，2015,32(5):651-659.QIAN Zhen,CAI Shibo,GU Yuqing,et al.Review of PV power generation prediction[J]. Journal of Mechanical&Electrical Engineering,2015,32(5):651-659.
• [5]LORENZ E,HURKA J,HEINEMANN D,et al.Irradiance forecasting for the power prediction of grid-connected photovoltaic systems[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2009,2(1):2-10.
• [6]王琦，季顺祥，钱子伟，等.基于熵理论和改进ELM的光伏发电功率预测[J].太阳能学报，2020,41(10):151-158.WANG Qi,JI Shunxiang,QIAN Ziwei,et al.Photovoltaic power prediction based on entropy theory and improved elm[J].Acta Energiae Solaris Sinica,2020,41(10):151-158.
• [7]刘家庆，张弘鹏，郭希海，等.基于SVR残差修正的光伏发电功率预测模型[J].电力工程技术，2020,39(5):146-151.LIU Jiaqing,ZHANG Hongpeng,GUO Xihai,et al.Prediction model of photovoltaic power generation based on SVR residual correction[J]. Electric Power Engineering Technology,2020,39(5):146-151.
• [8]李秉晨，于惠钧，刘靖宇.基于Kmeans和CEEMD-PELSTM的短期光伏发电功率预测[J].水电能源科学，2021,39(4):204-208.LI Bingchen,YU Huijun,LIU Jingyu.Prediction of shortterm photovoltaic power generation based on kmeans and CEEMD-PE-LSTM[J]. Water Resources and Power,2021,39(4):204-208.
• [9]刘国海，孙文卿，吴振飞，等.基于Attention-GRU的短期光伏发电功率预测[J].太阳能学报，2022,43(2):226-232.LIU Guohai,SUN Wenqing,WU Zhenfei,et al. Shortterm photovoltaic power forecasting based on attentiongru model[J].Acta Energiae Solaris Sinica,2022,43(2):226-232.
• [10]SALAMANIS A I,XANTHOPOULOU G,BEZAS N,et al.Benchmark comparison of analytical,data-based and hybrid models for multi-step short-term photovoltaic power generation forecasting[J].Energies,2020,13(22):5978.
• [11]LATEKO A A H,YANG H T,HUANG C M,et al.Stacking ensemble method with the RNN meta-learner for short-term PV power forecasting[J]. Energies,2021,14(16):4733.
• [12]LIU C C,LI M,YU Y J,et al.A review of multitemporal and multispatial scales photovoltaic forecasting methods[J].IEEE Access,2022,10:35073-35093.
• [13]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.GONG Yingfei,LU Zongxiang,QIAO Ying,et al. An overview of photovoltaic energy system output forecasting technology[J]. Automation of Electric Power Systems,2016,40(4):140-151.
• [14]王哲.基于ANN和时间周期性的光伏电站辐照度预测研究[D].北京：华北电力大学，2012.WANG Zhe.Solar irradiance prediction of PV power station based on artificial neural network and time-periodicity[D]. Beijing:North China Electric Power University,2012.
• [15]党存禄，武文成，李超锋，等.基于CatBoost算法的电力短期负荷预测研究[J].电气工程学报，2020,15(1):76-82.DANG Cunlu,WU Wencheng,LI Chaofeng,et al.Shortterm load forecasting based on CatBoost algorithm[J].Journal of Electrical Engineering,2020,15(1):76-82.
• [16]PROKHORENKOVA L,GUSEV G,VOROBEV A,et al. CatBoost:unbiased boosting with categorical features[C]//proceedings of the 32nd Conference on Neural Information Processing Systems(NIPS).Montreal,CANADA:[s.n.],2018.
• [17]党存禄，杨海兰，武文成.基于LSTM和CatBoost组合模型的短期负荷预测[J].电气工程学报，2021,16(3):62-69.DANG Cunlu,YANG Hailan,WU Wencheng.Short-term load forecasting based on LSTM and CatBoost combined model[J].Journal of Electrical Engineering,2021,16(3):62-69.
• [18]AKIBA T,SANO S,YANASE T,et al.Optuna:a nextgeneration hyperparameter optimization framework[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery&Data Mining.Anchorage,AK,USA.New York:ACM,2019:2623-2631.
• [19]王志宁.基于XGBoost的员工离职预测及特征分析模型[J].数字技术与应用，2021,39(3):193-196.WANG Zhining.Prediction and feature analysis model of employee resignation based on XGBoost[J].Digital Technology&Application,2021,39(3):193-196.
• [20]彭曙蓉，郑国栋，黄士峻，等.基于XGBoost算法融合多特征短期光伏发电量预测[J].电测与仪表，2020,57(24):76-83.PENG Shurong,ZHENG Guodong,HUANG Shijun,et al.Multiple-feature short-term photovoltaic generation forecasting based on XGBoost algorithm[J]. Electrical Measurement&Instrumentation,2020,57(24):76-83.
• [21]杨少华，李艾玲，雍浩，等.基于变量选择和DE-GWO算法的光伏发电功率预测[J].智慧电力，2018,46(12):38-44.YANG Shaohua,LI Ailing,YONG Hao,et al.Photovoltaic power generation forecasting based on variable selection and DE-GWO algorithm[J].Smart Power,2018,46(12):38-44.
• [22]杨修德.基于XGBoost的超短期负荷预测研究[D].银川：宁夏大学，2018.YANG Xiude.Study on ultra-short load forecasting based on XGBoost[D].Yinchuan:Ningxia University,2018.
• [23]陈垣毅.风能及光伏发电功率短期预测方法研究[D].杭州：浙江大学，2013.CHEN Yuanyi. Short-term prediction of wind power and photovoltaic power generation[D]. Hangzhou:Zhejiang University,2013.
• [24]邢峰，尹利，郑利宁，等.含高比例新能源及高载能用户的区域电网日前调度方法分析[J].内蒙古电力技术，2022,40(4):41-46.XING Feng,YIN Li,ZHENG Lining,et al. Analysis of day-ahead dispatching method of regional power grid with high proportion of new energy and high load consumers[J].Inner Mongolia Electric Power,2022,40(4):41-46.
• [25]孙星，梁瑞敏，张明.光伏电站监控系统应用方案研究分析[J].山西电力，2018(5):34-39.SUN Xing,LIANG Ruimin,ZHANG Ming.Research and analysis on application scheme of photovoltaic power plant monitoring system[J]. Shanxi Electric Power,2018(5):34-39.
• [26]毕会静，马燕峰，秦科源，等.不同配置模式下分布式光伏发电并网接纳仿真[J].电力电容器与无功补偿，2021,42(4):236-241.BI Huijing,MA Yanfeng,QIN Keyuan,et al.Acceptance simulation of distributed photovoltaic power generation grid-connection at different configuration modes[J].Power Capacitor&Reactive Power Compensation,2021,42(4):236-241.
• [27]孙辉，冷建伟.基于结合混沌纵横交叉的PSO-DBN的短期光伏功率预测[J].电测与仪表，2020,57(6):67-72.SUN Hui,LENG Jianwei. Short-term PV power prediction based on particle swarm optimization combined with chaos crossover for deep belief networks[J]. Electrical Measurement&Instrumentation,2020,57(6):67-72.
• [28]宋人杰，刘福盛，马冬梅，等.基于相似日和WNN的光伏发电功率超短期预测模型[J].电测与仪表，2017,54(7):75-80.SONG Renjie,LIU Fusheng,MA Dongmei,et al.A very short-term prediction model for photovoltaic power based on similar days and wavelet neural network[J].Electrical Measurement&Instrumentation,2017,54(7):75-80.
• [29]王士柏，王玥娇，王楠，等.基于小波分解与神经网络的光伏功率预测[J].山东电力技术，2021,48(5):1-7.WANG Shibo,WANG Yuejiao,WANG Nan,et al.Photovoltaic power prediction based on wavelet decomposition and neural network[J].Shandong Electric Power,2021,48(5):1-7.