Abstract: Considering the characteristic of Less-than-Container Load（LCL）shipping for small-batch orders overseas in manufacturing enterprises,the optimization of multimodal transport solutions for such orders was addressed.The mathematical model accounting for nodes including river,dry and sea ports was built by taking minimum total costs as object,and an improved adaptive genetic algorithm with an elite harmony strategy was designed.The proposed algorithm was first compared with Gurobi to validate its effectiveness.Then,it was compared with Stud GA,SGA and PSO algorithms,demonstrating superior performance in terms of solution quality,stability and time.An ablation study further analyzed the effects of the improvement strategies.Finally,a case study of a home appliance manufacturing enterprise verified the optimization effectiveness of the proposed method in real-world scenarios.

Key words: multi-modal transport, small-batch orders, less-than-container load, genetic algorithms

摘要： 针对制造企业海外小批量订单的拼箱海运需求,提出面向海外小批量订单的多式联运方案优化问题,考虑运输过程中涉及的内河港、陆港和海港3类节点,以所有订单的综合成本最小化为目标建立数学模型,设计一种基于精英和声策略的改进自适应遗传算法。在数值实验中,首先与Gurobi对比,验证算法的有效性;然后与种马遗传算法（Stud GA）、经典遗传算法（SGA）和粒子群算法（PSO）进行对比,结果表明算法在求解质量、稳定性和时间方面均具有优越性,并通过消融实验进一步对算法改进策略进行分析;最后,以某家电制造企业为例进行案例分析,验证优化方法在实际场景中的效果。

关键词: 多式联运, 小批量订单, 拼箱海运, 遗传算法