Volume 7: Urban Energy Systems: Building, Transport, Environment, Industry, and Integration

A New Modeling Method for Photovoltaic Modules Based on Extreme Learning Machine and I-V Curves Hui Yu, Zhicong Chen*, Lijun Wu, Qiao Zheng, Peijie Lin, Shuying Cheng



In this paper, a novel rapid modeling method is proposed for solar photo voltaic (PV) modules, which is based on extreme learning machine and current-voltage (I-V) characteristic curves. Firstly, original I-V curves are down sampled to reduce data redundancy, and a simple method is proposed to detect and remove abnormal I-V curves. Secondly, a single hidden layer feed forward neural network is proposed as the model, which is then trained by the extreme learning machine (ELM) algorithm. Finally, the proposed ELM based method is tested using a large data set of experimental I-V curves provided by the National Renewable Energy Laboratory (NREL). Experimental results show that the proposed ELM based method can shorten the modeling time to 0.2~0.4s, and the root mean square error (RMSE) can reach 0.0484%~0.374%. Compared with other conventional artificial neural network based methods,the proposed method can greatly shorten the modeling time and significantly improve the accuracy and the generalization performance of the modeling for PV modules.

Keywords PV modules, PV modeling, I-V characteristics, extreme learning machine

Copyright ©
Energy Proceedings