A Selective Maintenance Strategy on Network-Based System with Constrained Flow

The railway network, power system and pipeline system can be modeled as a network with many nodes and arcs with constrained flow. These critical infrastructures are designed to fulfill some continuous missions, such as traffic flow and electricity. However, due to aging and load-stress, the healthy state of the system will degrade and the reliability will decrease gradually. Different from the other systems which can be shut down for maintenance, the network-based system must provide continuous service meanwhile keep its reliability at a certain criterion. Thus, in this paper, a system is modeled by a network with single-origin and single-destination. The arcs degrade gradually and are modeled by a Gamma distribution with load-stress. The minimum flow of the network should be satisfied all the time. A selective maintenance strategy with sliding maintenance window is proposed based on the remaining useful life and the maximum change of network flow of each arcs. The aim of this study is to optimize the maintenance sequence of the arcs and meanwhile keep the continuous constraint of the network flow. Some numerical simulations will be conducted.