Green Path Optimization for E-Commerce Logistics under Carbon Neutrality Goals: A Multi-Objective Grey Wolf Optimizer (MOGWO) Approach

Abstract

In e-commerce logistics path planning, it is difficult to balance efficiency and environmental protection due to the lack of carbon emission considerations and multi-objective coordination. To address this problem, this paper constructs a green logistics path optimization method based on the MOGWO (MultiObjective Grey Wolf Optimizer) algorithm. This method first constructs a graph structure containing distribution centers, customer nodes, and path edges, and sets constraints such as transportation distance, time window, and vehicle capacity; then a linear carbon emission estimation model based on path length, vehicle type, and unit energy consumption parameters is established to provide input for multi-objective optimization. By unifying the dual objectives of minimizing transportation costs and minimizing carbon emissions, and introducing a time window violation penalty term, the path population is dynamically updated using the grey wolf hierarchical search mechanism of MOGWO, and the Pareto frontier is maintained in combination with non-dominated sorting to achieve target equilibrium. The experimental results show that the delivery costs of MOGWO standard orders, expedited orders, large orders, and overseas orders are 3.48, 4.98, 6.82, and 8.47 yuan per order, respectively, and the total transportation costs are 3482.56, 4998.72, 6815.34, and 8483.19 yuan, respectively, which are better than the traditional NSGA-II, PSO, and ACO, and the transportation cost control is efficient; the carbon emissions are reduced by 28.1%, 31.1%, 32.6%, and 32.9%, respectively, achieving effective control of carbon emissions. Based on MOGWO analysis, when the order delivery density is high and the vehicle loading rate increases from 40% to 90%, the unit carbon emission intensity decreases from 0.91kg/km to 0.649kg/km, and the total carbon emissions decrease from 1294.66kg to 899.54kg. The high loading rate can significantly improve carbon efficiency. The average route length in the urban distribution area is 12.84 km, and the average delivery time is 38.7 min, which is better than the 43.37 km and 96.5 min of the cross-regional distribution area, indicating that compact spatial structure and superior traffic conditions are important factors in achieving route optimization.