@Article{ee.2026.072751,
AUTHOR = {Wulfran Fendzi Mbasso, Idriss Dagal, Manish Kumar Singla, Muhammad Suhail Shaikh, Aseel Smerat, Abdullah Mohammed Al Fatais, Ali Saeed Almufih, Rabia Emhamed Al Mamlookol},
TITLE = {Drift-Aware Global Intelligent Optimization and Advanced Control of Photovoltaic MPPT under Complex Operating Conditions: A Cameroon Case Study},
JOURNAL = {Energy Engineering},
VOLUME = {123},
YEAR = {2026},
NUMBER = {4},
PAGES = {--},
URL = {http://www.techscience.com/energy/v123n4/66734},
ISSN = {1546-0118},
ABSTRACT = {Photovoltaic (PV) systems in the field operate under complex, uncertain conditions rapid irradiance ramps, partial shading, temperature swings, surface soiling, and weak-grid disturbances including off-nominal frequency and voltage distortion that degrade energy yield and power quality. We propose a drift-aware, power-quality-constrained MPPT framework that co-optimizes MPPT, PLL, and current-loop gains under stochastic frequency drift, while enforcing IEEE-519 limits (per-order I<sub>h</sub>/I<sub>L</sub> and TDD) during optimization. Unlike energy-only or THD-only methods, the design target integrates PQ constraints into the objective and is validated across calibrated drift scenarios with explicit per-order and TDD reporting. Operating scenarios are calibrated to Cameroon’s Southern Interconnected Grid and city-specific profiles (Douala/Yaoundé), combining measured-style irradiance/temperature traces, partial-shading patterns, and stochastic frequency drift up to ±0.8 Hz with synthetic contingencies. Across a 30-scenario campaign, the proposed controller achieves η<sub>MPPT</sub> = 99.3%–99.6% (vs. 98.6% Incremental Conductance and 97.8% Perturb-and-Observe), lowers DC-link ripple by 35%–48%, reduces oscillatory PCC power by ≈41%, maintains THD ≤ 2.5% (5% limit) and PF ≥ 0.99, and shortens irradiance-step settling from 85–110 ms to 50–65 ms. Sensitivity to PLL bandwidth shows a broad optimum (≈60–90 Hz) with minimum THD/ripple, and ablations confirm that explicit drift weighting is pivotal to ripple and THD suppression without sacrificing yield. The approach is controller-agnostic, firmware-deployable, and generalizes to other converter-interfaced renewables; we outline a short hardware-/HIL-validation path for adoption in Sub-Saharan grids.},
DOI = {10.32604/ee.2026.072751}
}