Maximum Power Point Tracking (MPPT) is critical for ensuring efficient energy extraction and voltage stability in standalone photovoltaic (SAPV) systems under partial shading and dynamic operating conditions. Conventional Perturb and Observe (P&O) methods suffer from slow convergence, steady-state oscillations, and sensitivity to noise. This paper proposes a sliding- inspired Online Convex Optimization (SOCO)-based enhanced P&O algorithm for adaptive duty ratio control in a PV-fed boost converter supplying a dynamic load. Two configurations are investigated: (i) constant irradiance with load switching, and (ii) variable irradiance with noise interference and dynamic loading. The proposed technique is validated through simulation using MATLAB/Simulink along with its implementation in real time using Software in the Loop(SIL) with OPAL-RT 4610. It is observed that the proposed technique exhibits overall power efficiency of 93.75%, thereby significantly higher than the traditional P&O algorithm that offers only 62.18%. The power tracking efficiency is found to be 98.48% under partial shading condition. The proposed method integrates online convex optimization with adaptive duty ratio control for real-time learning and improved robustness under partial shading, noise and load variations. Furthermore, the proposed method reduces the settling time from 0.18s (P&O) to 0.12s, corresponding to an approximate 33% reduction, demonstrating improved dynamic response under varying operating conditions.
Kurian et al. (2026) studied this question.