Performance Comparison of MPPT algorithm using P&O, INC and ANN Based Tracking

Introduction: The increasing demand for electricity underscores the necessity of renewable energy sources in enhancing electric grid stability, with solar and wind energy being prominent contributors. This study focuses on maximising solar power extraction through Maximum Power Point Tracking (MPPT) techniques. This research offers the comparative study of different MPPT algorithms with machine learning technique. To compare with machine learning technique this research has taken the prominent MPPT techniques of perturb and observe method and incremental conductance method. the above said two methods exhibits the higher volatile and oscillations in the peak point of power. To reduce the oscillations this research employs the higher adaptable Radial Basis Function (RBF) networks. These networks typically feature a single layer radial function, contributing to the versatility and adaptability of ANN in addressing complex tasks. The results are exhibiting the higher performance when compare to other techniques. Objectives: Compare the Performance Maximum Power Point Trancking algorithm using P&O, INC and ANN Based Mechanism. Conclusions: The simulation results highlight the superior performance of the ANN method in tracking the Maximum Power Point (MPP) under both rapidly changing and stable solar irradiation conditions. Its ability to quickly and accurately locate the MPP ensures optimal power extraction, outperforming the P&O and INC methods. While the P&O method demonstrates significant limitations, including poor performance under rapid irradiation changes and notable oscillations around the MPP under constant conditions, the INC algorithm shows moderate improvement. Although the INC method reduces oscillations compared to the P&O, it still incurs some power loss. Overall, the ANN approach emerges as the most reliable and efficient technique for maximizing power output in dynamic and steady solar environments.