Research on the MPPT algorithms of photovoltaic system based on PV neural network

In order to solve Maximum Power Point Tracking (MPPT) technology difficulties in photovoltaic system, 2-level neural network-genetic optimal algorithm is employed to estimate the photovoltaic battery model, taking into account possible influencing factors for battery output power such as light intensity, circumstance temperature, battery junction temperature, battery position. This method overcomes both power loss caused by oscillation around maximum power point with the traditional hill-climbing method and difficulty of training data with traditional neural network algorithm.