A Multi-objective Network Design Model for Post-disaster Transportation Network Management

Despite their inherent vulnerability to structural and functional degradation, transportation networks play a vital role in the aftermath of disasters by ensuring physical access to the affected communities and providing services according to the generated needs. In this setting of operational conditions and service needs which deviate from normal, a restructuring of network functions is deemed to be beneficial for overall network serviceability. In such context, this paper explores the planning of post-disaster operations on a network following a hazardous event on one of the network’s nodes. Lane reversal, demand regulation and path activation are applied to provide an optimally reconfigured network with reallocated demand, so that the network performance is maximized. The problem is formulated as a bi-level optimization model; the upper level determines the optimal network management strategy implementation scheme while the lower level assigns traffic on the network. Three performance indices are used for that purpose: the total network travel time (TNTT), the total network flow (TNF) and the special origin-destination pair (OD pair) accessibility. A genetic algorithm coupled with a traffic assignment process is used as a solution methodology. Application of the model on a real urban network proves the computational efficiency of the algorithm; the model systematically produces robust results of enhanced network performance, indicating its value as an operation planning tool.