Dynamic modeling of a compressed natural gas refueling station and multi-objective optimization via gray relational analysis method

Abstract

Compressed natural gas (CNG) refueling stations operate under highly dynamic thermodynamic conditions, requiring accurate modeling and optimization to ensure efficient performance. In this study, a dynamic simulation model of a CNG station was developed using MATLAB-SIMULINK, including detailed subsystems for multi-stage compression, cascade storage, and vehicle tank filling. Real gas effects were incorporated to improve prediction accuracy of the pressure, temperature, and mass flow rate variations during fast filling. The model was validated against experimental data, showing good agreement in both pressure rise and flow rate evolution. A two-stage multi-objective optimization approach was applied using Taguchi experimental design and gray relational analysis (GRA). In the first stage, storage pressures were optimized to maximize the number of vehicles filled and gas mass delivered, while minimizing compressor-specific work. The second stage focused on optimizing the volume distribution among the low, medium, and high-pressure tanks. The combined optimization led to a 12.33% reduction in compressor-specific energy consumption with minimal change in refueling throughput. These results highlight the critical influence of pressure levels and volume ratios in cascade storage systems on station performance. The presented methodology provides a systematic framework for the analysis and optimization of transient operating conditions in CNG infrastructure.