A bio-inspired algorithm for identification of critical components in the transportation networks

Critical components in a transportation or communication network are those which should be better protected or secured because their removal has a significant impact on the whole network. In such networks, they will be congested if they are being offered more traffic than it can process. In this paper, we employ principles of slime mould Physarum polycephalum foraging behaviour to identify the critical components in congested networks. When Physarum colonises a substrate, it develops a network of protoplasmic tube aimed at transporting nutrients and metabolites between distance parts of the cell. The protoplasmic network is continuously updating to minimize the transportation time, maximize the amount of cytoplasm pumped and minimize the overall length of the network. This optimization is achieved via a positive feedback between flux of cytoplasm and tube diameters. When a segment of a protoplasmic network is removed, the whole network reconfigures and thickness of tubes is updated till an equilibrium state is reached. The transient period from a disturbed state to an equilibrium state shows how critical the removed segment was. We develop a Physarum-inspired algorithm to identify critical links or nodes in a network by removing them from the network or calculating the transient period to new equilibrium state. The efficiency of the proposed method are demonstrated in numerical examples.