In a context marked by the increasing global energy demand and the urgency to adopt sustainable solutions, this research focuses on predicting solar isolation and optimizing photovoltaic (PV) systems using advanced machine learning techniques. We present here the results of an artificial intelligence (AI) model capable of predicting solar insolation over a given period while improving the efficiency of photovoltaic systems based on energy consumption. This model uses historical solar insolation data, as well as meteorological and environmental variables, to generate accurate forecasts of solar irradiation. The main objective of this research is to develop a robust predictive model for solar insolation while optimizing photovoltaic systems through machine learning techniques. We specifically aim to improve the accuracy of insolation forecasts over different time scales, whether hourly or daily, and to optimize the operational parameters of photovoltaic systems. This integrated approach will enhance the efficiency of photovoltaic systems and promote better integration of renewable energies into the electrical grid. To achieve these objectives, we will adopt a structured methodology in several steps, including data collection, modeling using machine learning algorithms, model validation through cross-validation techniques, and optimization of PV systems based on insolation forecasts. This innovative approach will not only allow for better planning of solar energy production but also make photovoltaic systems more efficient and suited to real needs, thus contributing to optimized management of energy resources.