目前常规的配送线路规划方法主要以配送开销或单位配送路线长度作为优化目标，并结合配送空间环境方面的限制对约束条件进行设计，由于忽略了配送过程中障碍物对配送效果的阻碍程度，导致算法的规划效果不佳。对此，提出基于模糊粒子群的多回程物流配送线路规划方法。首先，结合栅格法，根据障碍物信息对栅格编码，构建栅格编码与坐标之间的转换关系。其次，引入栅格危险度，对障碍物的阻碍程度表征。再次，以配送线路长度和配送危险度作为目标，构建出多目标函数，并结合配送设备的运行性能对约束条件设计。最后，利用粒子群算法，对构建出的物流配送线路规划函数迭代寻优，在约束条件下求解多目标函数，从而输出最优规划方案。在实验中，检验了该方法的规划效果。测试结果表明，采用该方法规划回程物流配送线路时，配送成本较低，具备较为理想的规划效果。

At present, conventional distribution route planning methods mainly focus on distribution costs or unit distribution route length as optimization objectives, and design constraints based on the constraints of distribution space environment. However, due to the neglect of obstacles in the distribution process, the planning effect of the algorithm is poor. A multi return logistics distribution route planning method based on fuzzy particle swarm optimization is proposed. Firstly, combining the grid method, the grid is encoded based on obstacle information, and the transformation relationship between the grid encoding and coordinates is constructed. Then, the grid hazard level is introduced to characterize the degree of obstruction of obstacles. Secondly, with the length of distribution routes and distribution risk as objectives, a multi-objective function is constructed, and constraint conditions are designed based on the operational performance of distributive equipment. Finally, using particle swarm optimization algorithm, the logistics distribution route planning function constructed is iteratively optimized, and multi-objective functions are solved under constraint conditions to output the optimal planning scheme. In the experiment, the planning effectiveness of the proposed method was tested. The test results show that when using the proposed method to plan the return logistics distribution route, the distribution cost is lower and it has an ideal planning effect.