THE MATHEURISTIC APPROACHES FOR HUB LOCATION OPTIMIZATION IN MULTIMODAL TRANSPORTATION NETWORKS

Abstract

This research addresses the integration of strategic, tactical, and operational planning within an intermodal hub distribution network. While previous studies have explored network problems with the aim of developing efficient solution methodologies, many rely on simplistic assumptions, often resulting in solutions based on limited variables that may not adequately reflect complexities. To bridge this gap, the study presents a comprehensive approach that incorporates realistic scenarios. Specifically, we formulate a multi-period, multi-allocation, and multi-capacity intermodal hub location problem using a mixed-integer nonlinear programming (MINLP) model with chance constraints. To effectively solve this complex model, a matheuristic algorithm—hybridizing metaheuristics (adaptive simulated annealing and tabu search) with nonlinear programming—is proposed. The algorithm demonstrates high efficiency, accurately solving small- to medium-sized problems while providing high-quality solutions for large-scale instances within limited computational time. Notably, the algorithm reduces average computational time by approximately 50% relative to the optimization approach. Empirical studies utilizing the algorithm examine the effects of variations in service quality and costs on the network structure and overall costs. The study reveals that reducing delivery frequency and increasing lead time influence costs and hub configurations, with economies of scale from diverse vehicle types lowering transportation expenses. However, expanding rail service often yields no benefits in this case, as factors like geography, costs, and operational constraints significantly impact the optimal logistics network design.