Forecasting Peak Supply of Solar PV Systems Utilizing Machine Learning Algorithms

Solar power generated from photovoltaic (PV) systems are the world's second-largest source of electricity followed by onshore wind and hydropower. Solar power is clean sustainable as its utilization does not emit greenhouse gases (GHG) into the atmosphere and it is continuously available so long as the sun continues to radiate light and heat on the Earth. However, a key factor which impacts the feasibility of solar power systems is the volatility involved in photovoltaic (PV) solar power generation. This volatility is caused by changes in weather and meteorological conditions. This study evaluates the effectiveness of the five regression Machine Learning Algorithms Linear Regression, Support Vector Machine, K-Nearest Neighbors, Random Forest, and Decision Tree in predicting the peak/maximum output power of the solar PV system across different seasons and geographical locations in South Africa. Historical meteorological and solar data for Pretoria, Johannesburg and Bloemfontein across four seasons was pre-processed and used for predicting the maximum output power of the solar PV system. The data was split into 80/20,80% for training and 20% for testing.