Sensitivity Analysis of Bi-Level Optimization for Resource Utilization in Maritime Container Terminals Under Disruptions.

Kurzfassung

Maritime container terminals are critical nodes in global supply chains, where disruptions and congestion can adversely affect operational efficiency and logistics performance. This paper presents a bilevel optimization framework for regulating container flows under perturbed operating conditions characterized by unusually high container arrival volumes. In particular, a disruption scenario is considered where over a time period of eight hours, every hour containers arrive at the terminal, compared to a standard scenario, in which containers arrive every twelve hours. The proposed model extends a convex container flow optimization problem by introducing a hierarchical decision structure for determining operational weighting parameters. The lower-level problem models the allocation of transport and handling resources within the terminal, while the upper-level problem selects parameter values to minimize congestion and improve container flow performance. The bilevel problem is reformulated using Karush–Kuhn–Tucker (KKT) conditions, and two solution approaches are investigated. The first approach incorporates a regularization term to improve numerical efficiency, whereas the second employs a Fischer–Burmeister-based relaxation of the complementarity constraints. The performance and solution quality of both methods are evaluated through sensitivity analyses with respect to input data and algorithmic parameters. Numerical results indicate that the relaxation-based approach is more robust to parameter and data variations than the regularization scheme. Moreover, it offers a practical tradeoff between computational efficiency and solution quality, making it suitable for congestion management in container terminals.