Comparative MPPT Performance in a PV System Using Different Neural Network Algorithms

In this paper, both feedforward neural network (FFNN) and adaptive network-based fuzzy inference system (ANFIS) are proposed to maximize the output power of a PV system with maximum power point tracking (MPPT) function in the DC-DC boost converter fed to a DC load. The proposed schemes are trained using practical irradiance and temperature data of a PV system. The performances of the proposed schemes are also compared with the one using traditional perturbation and observation (P&O) method. From the simulation outcomes, the MPPT performances of the studied PV system using the proposed FFNN and ANFIS are better than one using traditional P&O method.