Modeling a photovoltaic power system by CMAC-GBF

This paper proposed a photovoltaic (PV) power system model built by CMAC-GBF. Conventionally, the solar cell mathematic model is employed to form PV power system. But to build solar cell mathematic models, it needs solar cell manufacturers to provide detailed physical and manufacturing data. For researchers and users, it is not easy to access those data. Even when the data are obtained, their accuracy and reliability of environmental changes are all being doubted. In this project, CMAC is used to simulate PV systems, so the difficulty of accessing solar cell´s physical and manufacturing data can be ignored. Although the input range variations of PV systems are large, CMAC-GBF conquers the learning obstacles by using its excellent generalization ability, and outstanding performances are achieved. The data from a real PV system are employed to demonstrate the learning performance of CMAC.