考虑一对一取送约束的多车场零担物流配送路径优化

o address the challenges in the less-than-truckload (LTL) logistics distribution sector posed by multi-batch, small-volume freight demands and complex features like one-to-one pickup and delivery (P-D) and sequential pickup-before-delivery, the single-depot distribution model faces limitations in resource allocation, time fulfillment, and vehicle scheduling. This study investigated the multi-depot LTL distribution routing problem with one-to-one pickup and delivery constraints. A mixed 0-1 integer programming model was constructed with the objective of minimizing total logistics costs. A two-phase iterated greedy algorithm was designed, which generated the initial solution using a greedy nearest-neighbor heuristic strategy. Global search was performed through destruction-reconstruction operations, while local search incorporated multiple neighborhood operators. Simulation results demonstrate that the proposed two-phase iterated greedy algorithm outperforms exact solution methods, iterated local search, variable neighborhood search, and genetic algorithms in overall solution quality. Sensitivity analysis shows that setting a reasonable maximum vehicle service time significantly impacts distribution costs. Both excessively long or short times increase costs, necessitating a balanced value for cost minimization. Increasing the unit penalty cost for late arrivals raises total costs; thus, rationally reducing this coefficient effectively lowers total logistics distribution costs.