Experimental-Based Optimal Parameter Extraction for PEM Fuel Cell Semi-Empirical Model Using the Cloud Drift Optimization Algorithm

Accurate modeling of proton exchange membrane fuel cells (PEMFCs) is essential for predicting system performance under diverse operating conditions. This study introduces a refined semi-empirical modeling that combines experimental validation with an enhanced parameter estimation method based on the Cloud Drift Optimization (CDO) algorithm. The approach focuses on identifying seven key parameters of the nonlinear PEMFC model by minimizing the difference between experimentally measured and simulated cell voltages. To assess its effectiveness, the proposed CDO-based estimator was compared with several established metaheuristic algorithms, including the particle swarm optimizer and the tetragonula carbonaria optimization algorithm. The evaluation was performed using three commercial PEMFC stacks rated at 250 W, 500 W, and the NedStack PS6, as well as experimental data obtained from the Renewable Energy Laboratory at A’Sharqiyah University. Results demonstrate that the CDO algorithm consistently produced the lowest sum of squared errors (SSE) of 1.0337 and exhibited stable convergence across multiple independent runs with a standard deviation of 1.2114 × 10−7. Its reliable performance under both normal and degraded conditions confirms the algorithm’s robustness and adaptability, establishing CDO as an efficient and dependable technique for PEMFC modeling and parameter identification.