Enhancing Photovoltaic System Efficiency Through Neural Network-Based Tracking for Optimal Power Extraction

This study delves into the utilization of artificial neural networks (ANNs) to enhance the efficiency and stability of photovoltaic solar systems. Despite their clean and renewable energy source, photovoltaic systems encounter challenges arising from solar radiation, temperature variations, and environmental conditions, leading to fluctuations in current and voltage output and impacting power generation. The research addresses this concern by advocating for control strategies that optimize power extraction from the photovoltaic field. The central focus lies on the maximum power point (MPP), which denotes the optimal power transfer point on the current-voltage characteristic curve of a solar panel. Achieving precise MPP tracking is pivotal for bolstering system efficiency, given the task’s complexity in adapting to changing conditions. Existing tracking algorithms exhibit shortcomings in tracking rates and steady-state oscillations. To overcome these limitations, the study explores the application of ANNs in designing control algorithms. ANNs stand out for their agility in responding dynamically and adapting to nonlinear conditions. Yet, acquiring accurate training data for the controller remains a primary challenge. The investigation considers crucial factors like solar radiation, temperature, and optimal voltage as inputs for the controller. The proposed approach, built upon daily satellite-derived data for Latacunga city, yields promising results. It showcases an impressive average efficiency increase of up to 11.24%, alongside achieving rapid transient responses as swift as 0.56 ms. This research contributes to the advancement of photovoltaic technology by harnessing the potential of ANNs to revolutionize power extraction and utilization in solar systems.