| Summary: | Nowadays, artificial intelligent (AI) has growing rapidly with the technology advancement. One of the most popular section of AI is the machine learning (ML) which is applied widely for predictive analysis. Apart from that, the optimization algorithms are also from a section of AI are also popularly applied in system and cost minimization. Therefore, an accurate solar irradiance prediction is essential for optimizing photovoltaic (PV) system design to ensure efficiency and cost-efficiency. Throughout this project, Support Vector Regression (SVR) to predict solar irradiance based on historical weather data from a city is employed. Then, two powerful metaheuristic algorithms which are the Particle Swarm Optimization (PSO) and Differential Evolution (DE) are implemented to optimize the PV BESS system size and minimize the overall cost. The optimization process utilizes the predicted irradiance data with the exist data to determine the optimal configuration of the PV BESS system while considering the system size and cost. A comparative analysis between PSO and DE is conducted to evaluate their performance for achieving the most reliable and cost-effective solution. The results indicates that integrating SVR-based irradiance prediction with PSO and DE significantly enhances the PV BESS system design and further assist to reduce system costs and improve energy efficiency. This approach provides a useful structure for large scale solar system planning and decision making in PV energy applications.
|
|---|
