<p>Multi-objective problems (MOPs) can characterize problems in many real-world fields. The multi-objective grey wolf optimizer (MOGWO) is a highly effective method for solving these MOPs. However, the existing MOGWO still harbors notable defects. Firstly, the algorithm frequently succumbs to local optimization, leading to insufficient diversity in the final solution set. Secondly, the conventional crowding distance method can lead to uneven distribution of pareto front and may inadvertently exclude pivotal solutions, thereby undermining population diversity. Lastly, the position update method of MOGWO is relatively rudimentary and lacks adaptability, severely limiting its global search prowess. In response to these pressing issues, we introduce an enhancement to MOGWO by leveraging complex network theory. We meticulously construct a directed network that mirrors pareto dominance relationships, utilizing the influence of central nodes as a measure of vertical importance and the crowding distance to gauge horizontal diversity among individuals within the same dominance rank. By employing the topsis method, we derive a robust diversity ranking for individuals across both vertical and horizontal dimensions, thereby refining selection pressure control. Moreover, we incorporate quantum bit bloch to generate high-quality initial population. Additionally, we devise an improved differential evolution (DE) operator and a novel position update method, both aimed at bolstering global exploration capability and global optimality search. Drawing on these contributions, we propose the HSMOGWO. To validate its efficacy, we comprehensively compare HSMOGWO state-of-the-art multi-objective optimization algorithms on the ZDT, UF, and WFG test suites. Moreover, we apply HSMOGWO to address a suite of real-world problems. The experimental results unequivocally demonstrate that HSMOGWO exhibits exceptional competitiveness, surpassing the comparison methods in both solution quality and diversity.</p>