High-Accurate Parameter Identification of PEMFC Using Advanced Multi-Trial Vector Based Sine Cosine Meta-Heuristic Algorithm

Development and modeling of proton exchange membrane fuel cells (PEMFCs) need accurate identification of unknown factors affecting mathematical models. The trigonometric function-based sine cosine algorithm (SCA) may solve such problems, but it traps in local optima, making it inappropriate for larger optimization tasks. This paper introduces a novel multi-trial vector-based sine cosine algorithm (MTV-SCA) for the identification of seven unknown parameters of PEMFCs. The proposed MTV-SCA incorporates MTV methodology utilizing three control parameters to achieve the desired optimization targets. A key contribution of this work is the development of four distinct search strategies, to mitigating early convergence issues. These strategies leverage various sinusoidal and cosinusoidal factors to improve the algorithm. The optimization goal is to minimize sum square error (SSE) between measured and simulated stack voltages. Five PEMFC stack mode:250W, BCS 500W, SR-12, H-12, and Temasek 1 kW, validate the MTV-SCA algorithm’s efficacy and robustness Compared to previously established optimization approaches, MTV-SCA extracts optimum PEMFC parameters more accurately and reliably. Statistical testing further demonstrates the method’s durability and consistency. Analyzing PEMFC performance at different pressures and temperatures helps verify the adjusted parameters. Simulations show that the MTV-SCA solves difficult PEMFC parameter identification issues better than SCA and other approaches.