Multimodal Dynamic Freight Load Balancing

The urban traffic network has temporal and spatial characteristics whose changing conditions have often unpredictable effects on the flow of loads that include passengers and freight. As a result, the current traffic network is unbalanced, leading to high and low peaks of traffic in both time and space. The freight transportation chain can utilize these high and low peaks in the road and rail network in order to utilize more effectively available capacity. The purpose of this paper is to develop a coordinated multimodal dynamic freight load balancing (MDFLB) system to balance freight loads across the rail and road network. The MDFLB system collects and updates information from all the shipping companies and assigns freight loads to the available carriers using an optimization model while taking into account current and predicted dynamical changes in the associated networks. Since the freight loads can change the assumed states of the network, namely, the link travel times, which could then render the solution of the optimization problem no longer optimum, an iterative approach is considered involving online network simulation models. The simulation models are used to test and modify the optimization-based load balancing solution and estimate the new states of the network used by the optimizer. This feedback iterative approach guarantees that the overall cost function is non-increasing and it stops when it converges to a minimum or when a stopping criterion is satisfied depending on the time horizon of interest. A simulation case study that focuses on distribution of freight in an area that includes the two major sea ports in Southern California is used to demonstrate the effectiveness of the proposed coordinated MDFLB.