Parameter optimization study of porous media for enhanced heat transfer in liquid piston-type hydrogen compressor based on SOBP-SO algorithm

The liquid piston hydrogen compressor is considered the most promising compression technology for hydrogen refueling stations. However, the actual compression cycle is a polytropic process close to adiabatic, resulting in significant energy waste. To address this issue, this paper proposes a novel method for optimizing porous media parameters based on the advanced porous media-ionic liquid enhanced heat transfer scheme and the SOBP-SO algorithm. This method aims to balance the improvement in heat transfer efficiency provided by the porous media and its impact on fluid resistance, ultimately achieving the highest energy efficiency. The approach involves conducting numerous CFD simulations to generate a sufficiently dataset. The SO algorithm is then used to optimize the weights and thresholds in training BPNN, resulting in an SOBP neural network that can accurately predict heat dissipation efficiency and fluid resistance. These two SOBP networks are incorporated into a MATLAB model for energy analysis, enabling the determination the optimal porous media parameters. Simulations reveal that by using this porous media, the compression process achieves energy savings of 26.8 %. An experimental platform for the compressor has been constructed, and experimental tests have validated the effectiveness of the optimized porous media in enhancing heat transfer and achieving energy savings.