段震清;孙建民;梁凌;李庚达;崔青汝;伍权;

• 1:国家能源集团新能源技术研究院有限公司

摘要(Abstract)：

针对光伏阵列故障频发，且无法及时有效在线对故障类型进行识别的问题，提出了一种基于极度梯度提升(XGBoost)算法对光伏阵列进行故障诊断的方法。首先，基于Matlab Simulink仿真技术，建立光伏阵列仿真模型，针对正常、开路、短路、老化、阴影遮挡5种光伏阵列运行状态进行仿真，获取500例有效数据；其次，分析仿真数据特征变量之间的共线性关系，提取有效的特征变量作为模型的特征变量输入；然后，基于特征变量构建XGBoost故障诊断模型，并根据10折交叉验证方法优化超参数；最后，依据模型性能度量指标对XGBoost故障诊断模型的诊断结果进行评价，并分析模型特征变量的重要性。研究结果表明：基于XGBoost算法的光伏阵列故障诊断方法能简单、高效、实时在线对样本数据进行故障诊断，可应用于光伏阵列典型故障类型的有效识别。该故障诊断方法可为光伏电站现场运维人员提供技术支持，未来将在大型光伏电站，使用更大范围的实时数据开展使用和验证工作。

关键词(KeyWords)： 光伏阵列;故障诊断;XGBoost算法;特征变量

Abstract:

Keywords:

基金项目(Foundation): 无人值守光伏电站关键技术研究及应用(GJNY-20-123)

作者(Authors): 段震清;孙建民;梁凌;李庚达;崔青汝;伍权;

DOI: 10.19911/j.1003-0417.tyn20211114.01

参考文献(References)：

• [1]裴哲义,丁杰,李晨,等.分布式光伏并网问题分析与建议[J].中国电力,2018,51(10):80-87.
• [2]SPATARU S V,SERA D,KEREKES T,et al.Diagnostic method for photovoltaic systems based on light I-V measurements[J].Solar energy,2015(119):29-44.
• [3]SPATARU S V,SERA D,KEREKES T,et al.Monitoring and fault detection in photovoltaic systems based on inverter measured string I-V curves[C]//31st European Photovoltaic Solar Energy Conference and Exhibition,September 14-18,2015,Hamburg,Germany.[S.l.:s.n.],2015:1-10.
• [4]CHOUDER A,SILVESTRE S.Automatic supervision and fault detection of PV systems based on power lossesanalysis[J].Energy conversion and management,2010,51(10):1929-1937.
• [5]WANG W,LIU C F,CHUNG S H,et al.Fault diagnosis of photovoltaic panels using dynamic current voltage characteristics[J].IEEE transactions on power electronics,2015,31(2):1588-1599.
• [6]王培珍,郑诗程.基于红外图像的太阳能光伏阵列故障分析[J].太阳能学报,2010,31(2):197-202.
• [7]NIAN B,FU Z Z,WANG L,et al.Automatic detection of defects in solar modules:image processing in detecting[C]//2 010 6th International Conference on Wirele ss Communications Networking and Mobile Computing(WiCOM),September 23-25,2010,Chengdu,China.[S.l.:s.n.],2010:1-4.
• [8]DING H X,DING K,ZHANG J W,et al.Local outlier factor-based fault detection and evaluation of photovoltaic system[J].Solar energy,2018,164:139-148.
• [9]苏建徽,余世杰,赵为,等.硅太阳电池工程用数学模型[J].太阳能学报,2001,22(4):409-412.
• [10]陶彩霞,王旭,高锋阳.基于深度信念网络的光伏阵列故障诊断[J].中国电力,2019,52(12):105-112.
• [11]JARAMILLO S,MONTANE-MUNTANE M,GAMBUS P L,et al.Perioperative blood loss:estimation of blood volume loss or haemoglobin mass loss?[J].Blood transfusion,2020,18(1):20-29.
• [12]ALLEN M,POGGIALI D,WHITAKER K,et al.Raincloud plots:a multi-platform tool for robust data visualization[J].Wellcome open research,2019,4:63.
• [13]CHEN T Q,GUESTRIN C.XGBoost:a scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,August 13-17,2016,San Francisco,California,USA.[S.l.:s.n.],2016:785-794.
• [14]吴琼,余文铖,洪海生,等.基于XGBoost算法的配网台区低压跳闸概率预测[J].中国电力,2020,53(4):105-113.
• [15]RASCHKA S.Model evaluation,model selection,and algorithm selection in machine learning[EB/OL].(2018-11-13).https://arxiv.org/abs/1811.12808.