Power Management of a Solar System using Hybrid Maximum Power Point Tracking Algorithm

Abstract

The integration of artificial intelligence (AI) and data analytics is a focal point of this study. Machine learning algorithms and predictive models are employed to optimize panel orientation, predict solar irradiance, and detect system faults. By leveraging real-time data, AI-based approaches enable adaptive and proactive adjustments, resulting in increased energy output and prolonged panel lifespan.This paper presents a two-phase investigation into enhancing Maximum Power Point Tracking methods for solar systems. In the first phase, an Artificial Neural Network is trained using historical data from Incremental Conductance MPPT. The ANN leverages the existing knowledge of Incrimental conductance behavior to predict optimal reference voltages, enhancing tracking efficiency. The second phase involves integrating the trained Artificial Neural Network and Incrimental conductance to create a hybrid MPPT algorithm. This hybrid approach harnesses the ANN's adaptability and IncCond's rapid response and stability. Efficiency and stability assessments are conducted, evaluating power ripples, response time, oscillation frequency, and overall stability. Results highlight the hybrid MPPT's advantages, showcasing improved performance over individual methods. The paper contributes to advancing MPPT by introducing a novel hybrid approach that combines historical data-driven ANN training with Incrimental conductance benefits. This innovation holds the potential to substantially enhance solar energy harvesting efficiency and stability, impacting the system's overall output power.