Optimal Energy Utilization of Photovoltaic Systems Using the Non-Binary Genetic Algorithm

The article presents the energy model development based on the real-time output power of PV system installations at Kumamoto University, Japan. There are four sites of installation inside the university, which consist of two locations with thin-film silicon (5.76 kW and 7.56 kW) in the Faculty of Engineering and in the Faculty of Engineering Research Building, respectively, monocrystalline silicon (50 kW) in the Research Laboratory, and CIGS (30 kW) on the rooftop of the Forico Building. From the real-time output power information in one week continuous measurement during winter and summer in 2013, the accumulative energy and capacity factor are calculated. These data are then used as the raw information for developing an energy model in order to estimate the energy production based on capacity factor. The optimal capacity factor is obtained using the non-binary genetic algorithm (NB-GA). Obtaining the optimal capacity factor is important for the prediction of accumulative energy output following the forecasted meteorological information.