Volume 7: Proceedings of Applied Energy Symposium: CUE2019, China, 2019

A Novel Fault Diagnosis Method for PV Arrays Using Extreme Gradient Boosting Classifier Yutao Gan, Zhicong Chen*, Lijun Wu, Chao Long, Shuying Cheng, Peijie Lin


A new online fault diagnostic method for photo voltaic array is proposed in this paper, which is based on the Extreme Gradient Boosting (XGBoost)classifier. Firstly, the string current, array voltage, temperature and irradiance are measured by a monitoring system, from which a seven-dimensional fault feature vector is extracted as the input of the fault diagnosis model. Secondly, based on the XGBoost classifier, a new fault diagnosis model is established. Lastly, the feasibility and superiority of the proposed XGBoost based fault diagnosis model are tested by both Simulink based simulation and real fault experiments on a laboratory PV system. The correct rate of fault diagnosis in Simulink simulation is 99.99%, while the correct rate of fault diagnosis in laboratory PV power plant simulation is over 99.90%. Extreme learning machines (ELM) and Random Forests (RF) are tested for comparison. Experimental results demonstrate the superiority of the proposed XGBoost based model.

Keywords photo voltaic array, fault diagnosis, XG Boost, dynamic operating point

Copyright ©
Energy Proceedings