Using heuristic MCMC method for terminal location planning in intermodal transportation

摘要

In this paper, we consider the planning of terminal locations for intermodal transportation systems. With a given number of potential locations, we aim to find the most appropriate number of those as terminals to provide the economically most efficient operation when multiple service pairs are needed simultaneously. The problem also has an inherent task to determine the optimal route paths for each service pair. For this NP-hard problem, we present a Markov chain Monte Carlo (MCMC)-based two-layer method to find a suboptimal solution. In the lower layer, the routing for all service pairs given a particular location planning is solved through a table-based heuristic method that considers both efficiency and fairness. In the upper layer, by mapping the cost function into a stationary distribution, the optimal planning is solved based on a MCMC method that integrates advantages of both simulated annealing and slice sampling. Finally, the effectiveness of this heuristic MCMC-based method is demonstrated through computer experiments.