Optimization of the Distribution Paths of Agricultural Products in Cold Chain Logistics from a Resource Perspective

Abstract

Given the optimization problem in the distribution paths of agricultural products in cold chain logistics from a resource perspective, a deterministic model and a robust optimization model were established by comprehensively considering factors such as costs, product freshness, carbon emissions, and demands. Then, the models were solved by proposing a hybrid improved differential evolution with adaptive large neighborhood search (ALNS) algorithm. Simulation results demonstrate that under stable demands, the total cost of the deterministic model is 2419.95 yuan, the mileage is 204.6 km, and the vehicle full load rate reaches 90%. Under fluctuating demands, the total cost of the robust model increases by 6.76%, but the reliability of the scheme is guaranteed by reducing the load rate of some vehicles. The ALNS algorithm shows fast convergence in both models, which verifies its effectiveness in solving complex path problems. The proposed method provides differentiated decision support for cold chain logistic enterprises: The deterministic model is suitable for scenarios with accurate demand forecasting, while the robust model is more applicable to dealing with market fluctuations. In the future, the practicability of the models can be further enhanced by fusing dynamic traffic and multiperiod distribution constraints.