Enhancing Solar Power Output Predictions: Analyzing ARIMA and S-ARIMA Models for Short-Term Forecasting

As the world evolves towards sustainable energy, solar photovoltaic energy has become an important technology in the transition to renewable energy. Solar photovoltaic systems provide clean and more abundant energy by converting sunlight directly into electricity. However, interoperability of solar energy generation poses serious problems in terms of grid security and power management due to climate change, time of day and seasonal changes. Accurately predicting short-term solar photovoltaic energy generation is critical to ensuring efficient power distribution, reducing operating costs and ensuring product reliability in electricity usage. Therefore, statistics and machine learning have become important in solving these prediction problems.This paper explores the effectiveness of ARIMA (AutoRegressive Integrated Moving Average) and SARIMA (Seasonal AutoRegressive Integrated Moving Average) models in short-term hourly forecasting of solar PV power. The study evaluates the models’ performance in predicting hourly data, emphasizing its applicability to weather prediction, energy consumption and financial markets. Comparative analysis is conducted to highlight the strengths and limitations of each model.