Performance evaluation of Maximum Power Point Tracking (MPPT) algorithms in photovoltaic systems and wind turbines under various conditions

Optimal renewable energy harvesting within short times is critical for hybrid off-grid systems, where variability directly affects reliability. This study evaluates eight MPPT methods which are Perturb and Observe (P&O), adaptive P&O, incremental conductance (INC), adaptive INC, Particle Swarm Optimization (PSO), PSO with Proportional–Integral (PI) Controller, Artificial Neural Network (ANN) and ANN with PI. MATLAB/Simulink simulations have been conducted for solar PV and wind turbines. For solar PV, algorithms have been tested under deep low-start irradiance temperature stress at 50 W/m2 and 0 °C, rapid irradiance steps, ramp–flicker profiles, full-sun to shade transitions, and irradiance spikes with high-temperature stress at 45 °C. Wind algorithms have been tested under sudden wind-speed changes, including sub-rated conditions at 16 m/s. Performance metrics have included efficiency, convergence time, robustness, energy yield, and computational cost. Adaptive and intelligent methods have consistently outperformed conventional MPPT. For PV, Adaptive INC has achieved 94.14% efficiency, 100% robustness, and minimal oscillation. ANN with PI has delivered similar results but has required less computational time and has offered higher stability. For wind, PSO with PI has been the fastest (≤2 s), fully reliable with 100% success, and has yielded 6% more energy than standard PSO. Overall, combining metaheuristic and neural controllers with PI tuning improves dynamic performance, stability, and energy harvesting, providing a strong reference for MPPT selection in autonomous renewable systems. Copyright © 2026 Praise Worthy Prize S.r.l.-All rights reserved.

Actions (login required)

View Item