325 / 2024-04-24 14:16:57

Resource allocation in air-rail-integrated co-modality under both demand and supply uncertainties

air-rail-integrated co-modality,mixed-integer stochastic programming,Sample Average Approximation,real-time algorithm

The rapid expansion of cross-border e-commerce has promoted the demand for transporting high-value and time-sensitive goods. Air-rail-integrated co-modality emerges as a promising solution to leverage under-utilized capacity for cargo delivery. This novel co-modality offers additional revenue and promotes the development of eco-friendly transportation, This study tackles the resource allocation challenges in the co-modality. Cargo demand can be fulfilled via dedicated cargo aircraft or co-modality. In the latter case, the resource allocation decisions involve selecting appropriate trains and aircraft for cargo transport. To ensure efficiency and flexibility on the operational level, the air-rail co-modality operator needs to consider critical factors such as transshipment synchronization and uncertainty on the supply and demand side. To formulate this problem, a three-stage mixed-integer stochastic model is proposed to minimize the total costs associated with transportation, holding, transshipment, delays, and ad-hoc options. We consider the Hong Kong West Kowloon Station and the Hong Kong Airport as two transshipment hubs to conduct the numerical experiments. We apply the Sample Average Approximation to solve the approximation model using the Gurobi solver. Preliminary results show that medium-scale instances with 100 scenarios, encompassing 40 origin-destination pairs, 20 trains, and 20 passenger aircraft, can be resolved within tens of seconds. We are developing heuristic algorithms for large-scale instances to obtain accurate and fast solutions. Furthermore, we introduce a rolling horizon framework for real-time optimization of the resource allocation plan, accommodating the uncertainties of demand and supply.