Urban agglomeration low-carbon logistics network design with stochastic demand

Aligning with the promotion of green development in the urban agglomeration, this study investigates the low-carbon design of multi-stakeholder multi-modal logistics network with uncertain future demand. Given a limited budget for the construction of logistics parks, the goal of the logistics authority is to design a network configuration with the least expected CO2 equivalent (CO2e) emissions, including capacities of logistics parks and subsidies for rail links, while allowing the logistics carriers to choose the routes with the lowest generalized costs in the designed network structure. This problem is formulated as a novel two-stage stochastic bi-level programming (TSSBLP) model, in which the second stage decision involves a bi-level model with uncertain demand. In order to effectively solve this problem, an exact algorithm based on progressive hedging embedded with simplicial decomposition method (PH-SDM) is presented to obtain the optimal solution of the proposed model. Finally, a case study is applied to validate the effectiveness of proposed models and PH-SDM. Numerical results show that the subsidy strategy is an effective measure to reduce carbon emissions, and the economic subsidy rate should be set according to the realistic target.