Optimal design of logistics network for remanufacturing under uncertainty

Remanufacturing logistics has obvious uncertainty, such as uncertainty of timing and quantity of returned products, balancing between returns of used products with demand for remanufactured products and materials recovered from returned items. The return of used products, demand for remanufactured products and recovery of used products are taken as random parameters, and a stochastic programming model is proposed to minimize the costs of logistics network considering multi-product, multi-cycle, capacity constraints and other factors. The numbers and locations of various logistics facilities and the volume of corresponding products are determined through the established model. A hybrid genetic algorithm integrating stochastic stimulation and linear programming is submitted to efficiently solve the model, which improves the capacity of local optimization. Computation results show that the model can better meet the requirements of remanufacturing companies in strategic decision-making because it takes into account more actual factors compared with the deterministic models.