Multi-objective particle swarm optimization algorithm based on decomposition and hypervolume for synthesis gas production

Abstract

Synthesis gas has a very wide range of applications in the chemical industry. In this paper, the experiment is needed to achieve the highest CH4 conversion rate, the highest CO selectivity, and the lowest H2/CO ratio in synthesis gas production by controlling the feedstock O2/CH4 feed ratio, gas hourly space velocity, and temperature in the synthesis gas production process. Multiple objectives can conflict with each other, so multi-objective optimizations are needed to achieve a balanced result for all parties. We propose a new multi-objective particle swarm optimization algorithm based on decomposition and hypervolume (MPSO/DH). This algorithm uses decomposition to divide the objective space into several sub-regions and selects the optimal solution in each sub-region, then classifies the remaining solutions and selects the solutions that can make the maximum hypervolume in sub-regions, to ensure that the hypervolume of the next generation population is also as large as possible, and to achieve the goal of improving diversity and convergence at the same time. Finally, we applied this algorithm to synthesis gas production and verified the effectiveness of the algorithm.