Multi-objective Cold Chain Logistics Path Optimization Using Heu-ristically-Initialized and Catastrophe-Enhanced NSGA-II for E-com-merce Distribution

Abstract

The swift advancement of e-commerce poses challenges to the cost, efficiency, and quality assurance of fresh food and pharmaceutical Cold Chain Logistics (CCL). To reduce transportation costs, decrease refrigeration energy consumption, improve customer satisfaction with time and freshness of goods, the research proposes an e-commerce CCL distribution path optimization model based on improved non-dominated sorting genetic algorithm II. This model takes the total transportation cost, customer time sat-isfaction, average freshness of goods, and total refrigeration energy consumption as multiple objectives. It generates high-quality initial solutions through heuristic population initialization and combines dy-namic disaster mechanisms to avoid local optima, enhancing the algorithm's global search ability and convergence speed. In the experimental verification of Solomon Benchmark Problem and Berlin52 stand-ard dataset, a population size of 100 and a maximum iteration of 300 are set up in the experimental environment. The proposed optimization method is compared with the original algorithm, multi-objective evolutionary algorithm based on adaptive reference points, and non-dominated sorting genetic algorithm II based on simulated annealing improvement. Experiments show that the optimization model is better than the traditional algorithm in indicators such as total transportation cost, refrigeration energy con-sumption, customer time satisfaction, and product freshness. Among them, the total transportation cost is the lowest at 5923.47 yuan, and customer time satisfaction and average product freshness reach 0.954 and 0.962 respectively. The total refrigeration energy consumption drops to 87.93 kWh, the optimal route mileage is 436.59 km, the delivery time is 945.38 minutes, and the cargo damage rate is 2.35%. The results show that this optimization method can efficiently coordinate multi-objective conflicts, achieve path opti-mization, cost reduction, and service quality improvement, and provide stable and efficient decision sup-port for e-commerce CCL distribution.