Abstract

The increasing interest in renewable energy and the decreasing costs of solar panels position solar electricity favorably for widespread adoption. However, the rapid integration of intermittent renewable energy sources brings challenges and the risk of power instability between available power generation and load demand. Therefore, precise solar Photovoltaic (PV) power forecasting is crucial for maintaining system reliability and maximizing the integration of renewable energy. Current solar PV power forecasting methods are vital tools for ensuring system reliability and optimizing renewable energy integration. This paper offers a thorough and comparative review of existing Machine Learning (ML) based approaches to PV power forecasting, particularly for short-term horizons. We present an overview of the factors influencing solar PV power forecasting and summarize the existing forecasting methods in the literature, with an emphasis on ML-based models. To enhance the comparison and provide deeper insights into advancements in this area, we simulate the performance of various ML methods used in solar PV power forecasting and discuss the results.