科研之友
微信
新浪微博
Facebook
分享链接
摘要
To increase the energy yield of installed solar photovoltaic (PV) systems, proper operation of all components should be ensured. One way to detect any misbehavior is to estimate the electric power of a PV system using data from meteorological stations or an in-situ integrated weather station to calculate the electric power of a PV system using a PV module performance mathematical model. By comparing the calculated and the actual electric power, the performance of a PV system can be determined. Such a PV system monitoring greatly depends on accuracy of measured weather parameters and precision of a PV module performance model used. Simple heuristic models are easy to implement but lack accuracy in particular under lower irradiances, higher share of diffuse light or higher solar incident angles. To improve accuracy, we developed a new PV module performance model with splitting diffuse and direct light including solar incident angle. As a consequence, the new model requires additional measurement of diffuse irradiance, but simplicity is maintained using simple equations with low number of fitting coefficients that were extracted for a typical multi-crystalline (mc-Si) and thin film copper indium gallium selenide (CIGS) PV module, based on outdoor monitoring data in Ljubljana, Slovenia. The new model was compared to two other heuristic models and a decrease of root mean square error (RMSE) in eight snowless month period (during October 2013 to September 2014) was observed, on average by 1%. Much larger improvement with the new model is obtained at lower irradiances: RMSE at irradiances of 200 W/m(2) in June 2014 was three times lower compared to the other two models. The new PV module performance model with separated diffuse and direct irradiance component assures higher accuracy of prediction and thus higher reliability of PV system fault detection applications.
- 出版日期2015-3
