Performance Analysis of Maximum Power Point Tracking Algorithms for PV Systems Under Partial Shading Conditions in North West Nigeria

This study analytically and experimentally compares the performance of maximum power point tracking algorithms for photovoltaic systems operating under partial shading conditions typical of North West Nigeria, where Harmattan dust, irregular tree shading near buildings, and bird perching create non-uniform irradiance across PV arrays. Under partial shading, the PV array power-voltage characteristic exhibits multiple local power peaks, and conventional MPPT algorithms such as perturb and observe and incremental conductance fail to reliably locate the global maximum power point, reducing energy yield. Six MPPT algorithms are evaluated: perturb and observe, incremental conductance, particle swarm optimisation, grey wolf optimisation, modified firefly algorithm, and a hybrid PSO-gradient descent approach. Each algorithm is implemented in MATLAB and evaluated on a validated PV array model constructed using the single-diode equivalent circuit, with partial shading patterns derived from field measurements at two PV installations in Kano and Sokoto. Evaluation metrics include tracking efficiency, convergence speed, and steady-state oscillation magnitude. Laboratory hardware validation was conducted using a dSPACE rapid prototyping platform with a programmable DC power supply emulating shaded PV characteristics. Results demonstrate that swarm-intelligence algorithms consistently locate the global MPPT under all tested partial shading patterns with efficiencies above 99 percent, while perturb and observe achieves the global peak in only 43 percent of cases. Grey wolf optimisation shows the best balance of tracking efficiency and convergence speed under the tested conditions. Keywords: MPPT, partial shading, photovoltaic, metaheuristic optimisation, North West Nigeria.