Modelling, Simulation and Genetic Algorithm Optimization of a Photovoltaic System Validated with Field Data from Lagos, Nigeria

Reliable prediction of photovoltaic performance under variable tropical conditions remains a barrier to the cost-effective deployment of solar energy in sub-Saharan Africa. This study develops, validates, and optimizes a physics-based model of a photovoltaic system using MATLAB/Simulink and the Simscape Electrical toolbox. A single-diode equivalent circuit model was constructed from the governing photovoltaic cell equations and parameterized with real operational data collected from a mini solar installation in Lagos, Nigeria, comprising twelve 0.5 V cells connected in series to deliver 6 V and a peak power of 6 W. The model simulated module output power in response to changes in irradiance, ambient temperature, and load, and its current-voltage and power-voltage characteristics were compared against measured field data, showing strong agreement across operating conditions. A genetic algorithm was then applied to optimize panel configuration and system sizing with the objectives of maximizing annual energy output and minimizing the levelized cost of energy. The optimized configuration achieved a 10 percent increase in energy output, a 5 percent reduction in levelized cost of energy, and a 2 percent improvement in system efficiency. Sensitivity analysis identified solar irradiance and ambient temperature as the dominant performance drivers: a 10 percent increase in irradiance raised energy output by 12 percent, while a 5-degree Celsius temperature rise reduced output by 3 percent. The optimized system is projected to generate 3567.57 kWh annually at a levelized cost of NGN 83 per kWh with a payback period of seven years. The framework offers a transferable, low-cost pathway for designing efficient residential and commercial photovoltaic systems in high-irradiance developing regions.