Assessment of a modified Coronavirus Disease Optimization Algorithm for Parameter Estimation of Proton Exchange Membrane Fuel Cells

Abstract

Parameter identification of a proton exchange membrane fuel cell (PEMFC) involves estimating the unknown model parameters required to build an accurate predictive representation of fuel-cell performance using optimization-based techniques. Because these parameters are often unavailable in manufacturer datasheets, their estimation is crucial for reliable performance prediction and system evaluation. In this work, a memory-based Coronavirus Disease Optimization Algorithm (mCOVIDOA) is proposed as an enhanced optimization method for extracting PEMFC parameters. Six unknown model parameters are identified and compared using six optimization approaches: mCOVIDOA, the standard Coronavirus Disease Optimization Algorithm (COVIDOA), Tunicate Swarm Algorithm (TSA), Grey Wolf Optimizer (GWO), Chimp Optimization Algorithm (ChOA), and Moth-Flame Optimizer (MFO). During the optimization process, these parameters are treated as decision variables, and the objective is to minimize the sum of squared errors (SSE) between simulated and measured cell voltages. The results show that mCOVIDOA achieves a lower SSE (1.9454) and consistently outperforms the original COVIDOA as well as GWO, TSA, ChOA, and MFO in terms of convergence quality and solution accuracy. Owing to its accurate prediction capability and faster convergence, mCOVIDOA shows strong potential for digital-twin development of fuel-cell systems and advanced control applications in automotive energy systems.