Research on Multi-Objective Optimization of Low-Carbon Cold Chain Routes Using LNS-NSGA2 Algorithm

This paper addresses the high costs, stringent time sensitivity, and perishability issues encountered in the cold chain logistics of fresh products. To tackle these challenges, a multi-objective low-carbon cold chain route optimization model is established. This model aims to simultaneously minimize the total cost (comprising transportation cost, cargo damage cost, and carbon emission cost), while maximizing customer satisfaction (encompassing both time satisfaction and quality satisfaction). To achieve this goal, we propose the LNS-NSGA2 algorithm, which integrates the advantages of local search strategies with genetic algorithms. Building upon the traditional NSGA2 algorithm, the LNS-NSGA2 algorithm incorporates the “destroy” and “repair” mechanisms from the LNS (Large Neighborhood Search) algorithm, coupled with a binary tournament selection strategy, to enhance the algorithm's local search capability and thus exploring the solution space more efficiently. Through a comparative analysis of a series of large-scale experimental cases, the results demonstrate that compared to the traditional NSGA2 algorithm, the LNS-NSGA2 algorithm effectively reduces distribution costs while significantly improving customer satisfaction. It can notably enhance the quality of the Pareto Frontier, providing superior distribution schemes for fresh food trading outlets and cold chain logistics enterprises.