Component Importance Balancing of Multi-Commodity Networks

Abstract

Determining which network components to focus on in case of a disruption can be determined by the component with the greatest importance to the network. This study extends a balancing component added to a list of optimization targets in multi-state system to a multi-commodity network. A bi-objective optimization problem is developed in which one objective seeks to minimize the difference in a flow-based component importance measure across one-at-a-time interdiction scenarios by adding capacity to components, while the other minimizes unmet demand to encourage flow through the network. The optimization approach is applied to a multi-commodity Swedish railway network, where a set of link capacity increases per commodity are suggested based on the desire to balance the network.