Data-driven assisted multiobjective routing optimization for multimodal transportation under fuzzy demand

Multimodal transportation, as an efficient and comprehensive transport mode in the modern logistics system, significantly improves logistics efficiency and reduces costs by integrating various transportation modes. However, due to uncertainties such as weather conditions, logistics companies cannot accurately grasp and predict freight transportation demand. Additionally, the time-sensitivity requirements of multimodal transportation frequently lead to delays in freight transportation. To characterize the randomness and uncertainty of transportation demand parameters, triangular fuzzy numbers are introduced as a descriptive tool. A multiobjective optimization model for low-carbon multimodal transport routing under uncertain demand is constructed, aiming to minimize total transportation cost and carbon emissions. Monte Carlo simulation combined with data-driven methods is employed to sample and analyze uncertain demand, effectively addressing demand uncertainty. To solve the developed model, a novel multi-form competitive swarm optimization algorithm is proposed, which enhances both the quality and efficiency of the solutions. The experimental results indicate that compared to peer algorithms, the proposed algorithm achieves a better tradeoff in diversity, convergence, and feasibility, ultimately obtaining superior performance.