Adaptive Neuro-Fuzzy Approach for Forecasting of Solar Power Generation

Abstract

This paper highlights the prediction of power generation from a solar panel with the help of the ANFIS. Solar irradiance and atmospheric temperature in a place are not even throughout the year. It depends on geographical position (i.e., longitude and latitude) and the environmental situation of the place. Power generation from a PV panel is highly subjective by solar irradiance and temperature. Pollution, fog, smoke dust cloud, etc. may also affect these two. So, changes in environmental conditions may affect the prediction of power generation from solar PV panel. In this paper, power generation is predicted for different irradiances and temperatures with the help of the ANFIS. ANFIS results are validated with the real-time experimental value which gives an improvement in predictive accuracy. This also signifies the ANFIS capability to estimate power in a different situation which could be helpful for maximum power point tracking.