A comprehensive comparative study of three MPPT controllers for photovoltaic systems based on tracking performance and steady-state ripple indicators

Introduction. The efficiency of photovoltaic (PV) energy conversion is fundamentally dependent on the accuracy and stability of maximum power point tracking (MPPT) algorithms. Problem. Most comparative studies of MPPT algorithms focus primarily on evaluating tracking efficiency (h) and convergence speed (tr), largely overlooking the impact of steady-state electrical ripples on converter component stress. This represents a critical gap in reliability-oriented PV system design. Goal. To compare three MPPT strategies: perturb and observe (P&O), fuzzy logic control (FLC), and artificial neural network (ANN)-based control coupled with an anti-windup PI regulator, using six performance indicators. Methodology. Three test scenarios were conducted through MATLAB/Simulink simulation of a PV system under varying irradiance and temperature conditions. To track the MPP, the FLC uses the power gradient (E) and its variation (ΔE) as inputs to a fuzzy inference engine with adaptive step output, while the P&O perturbs the duty cycle by a fixed step ΔD, the ANN+PI employs a feedforward architecture trained to predict the optimal voltage reference Vmpp*, the error e(t) is tracked by an anti-windup PI regulator to adjust the duty cycle continuously. Results. The ANN+PI achieves the fastest convergence and the lowest ripple across all metrics (tr =11 ms, ΔP= 0.85 W, ΔV = 0.837 V, ΔI = 0.163 A), confirming the effectiveness of closed-loop voltage regulation in minimizing component stress. The FLC achieves the highest average tracking efficiency (ηavg= 99.22 %), demonstrating the energy extraction advantage of adaptive step adjustment. Scientific novelty. Unlike studies mainly focused on tracking efficiency and dynamic response, this study uses a quantitative indicator, notably power ripple (ΔP), voltage ripple (ΔV) and current ripple (ΔI), as indirect measures of the electrical stress and operational stability of the converter components. Practical value. These findings can be directly exploited in real PV system design to select the most suitable MPPT strategy. References 20, tables 5, figures 17.