Abstract To address the challenges of high costs, volatility, and multiple constraints in cross-border e-commerce logistics of home appliances from Japan to East China, this study proposes a two-segment chromosome genetic algorithm (GA) optimization mechanism to systematically compare ten logistics operation models that are currently feasible for enterprises. The first segment encodes the suitability of candidate logistics models, while the second segment simultaneously evolves the weight coefficient of key performance indicators. A global search is conducted in a multi-objective space using fitness-proportional selection, BLX- α crossover, and adaptive Gaussian mutation through which both the suitability and the coefficient weights are synchronously adjusted during the search process to obtain a comprehensive evaluation that is more consistent with real business scenarios. The experiment is based on company-collected operational data and industry survey results, incorporating transportation, warehousing, return logistics, and carbon emissions. The GA results indicate that the Off-peak Transit Time is the most critical performance indicator, leading to a strong preference for air shipping, with approximately 60% of orders allocated to air consolidation. As reason, the time and inventory costs associated with short-sea routes weaken the traditional advantages of sea shipping, while the economies of scale of consolidation offset the typically high cost of air shipping. The study demonstrates the applicability of an enhanced GA in medium-range cross-border logistics system design, offering practical decision support for enterprises seeking a balanced trade-off among cost efficiency, lead time, service quality, and sustainability.
Huang et al. (Fri,) studied this question.