Anticipating Solar Photovoltaic Panel Output One Day in Advance with Gaussian Mixture Model

The escalating demand for energy across various applications has led to the depletion of non-renewable energy sources. As an alternative, Solar Photovoltaic (PV) has gained prominence in meeting energy requirements. Forecasting the output power of PV panels has become essential to optimize solar energy utilization. Solar power generation is pivotal in innovative grid applications for efficient energy distribution. However, challenges exist in ensuring the stability and effective energy management of innovative grid systems. This paper presents an approach for predicting the output power of PV panels by utilizing past solar power output and input param-eters. To forecast solar irradiation and operating temperature for the next day, the Gaussian signal fitting method is employed. The Linear Regression model, a supervised machine-learning technique, is applied to establish a predictive relationship between input and output. Furthermore, the Gaussian Mixture Model (GMM) is utilized to estimate the output power of PV panels.