Hybrid partheno-genetic algorithm for multi-depot perishable food delivery problem with mixed time windows

Energy cost for maintaining the freshness of food and the cost for timely delivery are two main features to consider in food delivery. Stale food and early or late delivery reduce customer satisfaction and are thereby detrimental to corporate profits. To avoid this problem, food companies commonly use refrigerated trucks and multiple distribution centers, which involve additional costs. As such, planning a reasonable path to retain customer satisfaction while reducing as much costs as possible during the operation is of considerable importance. This feature varies according to transportation and services. Therefore, this study focuses on and models food delivery in cold chain logistics as a multi-depot vehicle routing problem with mixed time windows (MDVRPMTW), which is a complicated nonlinear optimization based on a group of complex objectives and constraints. Clustering and sorting method are used for the initialization of population to reduce complexity. Subsequently, an improved hybrid partheno-genetic algorithm (HPGA) is proposed to solve this problem. The algorithm uses gene block-based crossover and mutation operations and local elite strategy to adjust customer assignment and improve performance. Computational results of the benchmarks and MDVRPMTW indicate the effectiveness of the proposed algorithm. Furthermore, a real case study is carried out to validate the feasibility of the proposed model.