Cold chain logistics route optimization for fresh agricultural products based on time-varying traffic and HHO algorithm

With the ongoing growth of urban cold chain logistics distribution demand, route planning faces severe challenges in dealing with time-varying traffic, controlling cargo damage, and ensuring the quality of fresh produce. To this end, a cold chain logistics route optimization model integrating time-varying traffic factors and freshness decay mechanism is constructed, and an improved Harris Eagle optimization algorithm combining Circle mapping initialization, Cauchy mutation, and adaptive weight strategy is proposed. The algorithm performance is validated based on the classic vehicle path planning benchmark dataset Solomon public dataset to ensure the transparency and comparability of experimental results. The ablation experiment outcomes reveal that the introduced algorithm outperforms the original algorithm in regard to optimal value, mean value, and standard deviation. Specifically, the average distribution cost is reduced to 16,497.6 yuan, the standard deviation is controlled at 118.7, and the convergence stability is significantly improved. The comparative experiment results show that the loss values on the training set and test set are the lowest, converging to 0.03 and 0.04, while the running time is the shortest, only 24.6 seconds in the testing phase. In real-world cold chain distribution scenarios, the proposed method controls the cargo damage rate at 7.1%, 4.8%, and 7.9% during morning peak, flat peak, and evening peak periods, respectively, with an average freshness maintained above 0.918. The empirical outcomes further demonstrate that the introduced approach has superior route optimization capability in complex dynamic environments, providing a practical solution for intelligent scheduling of cold chain logistics.