Supply-demand interaction is a challenge that must be considered in no-notice evacuation modeling. It is quite common that during an evacuation event, demand is much greater than supply for an extended time period, resulting in severe and prolonged traffic congestion. The severity and temporal extent of such congestion cannot be estimated by simple calculation because of the traffic flow phenomenon in which the traffic throughput is much less than the nominal capacity under severe congestion. Even more, evacuees change their decisions before or during evacuation in response to various traffic management strategies, such as radio and message sign information or contraflow lanes and improved signal control. Once the traffic management decisions are modified, the traffic demand on various evacuation routes would change, as does the congestion resulting from this new demand-supply interaction.The system performance of such an interaction can be properly captured through a descriptive traffic simulation model combined with a prescriptive optimization model for the evacuation strategies. The selection and implementation of no-notice evacuation strategies need to be driven by one or several system objectives. In the prescriptive concept, the best possible strategy is then established through a system optimization approach.Most traditional prescriptive models use only static flow without considering vehicular flow dynamics, such as shock wave propagation or queue formation and dissipation. In recent years, the cell transmission model (CTM) (1) has been used as part of the system-optimal (SO) dynamic traffic assignment (DTA) method (2) in several evacuation studies (3,4), to determine the multidimensional optimal decisions simultaneously, including evacuees' departure time, destination, and route choices (5).The "information" component is another important element that needs to be properly captured in evacuation modeling. Chiu and Mirchandani studied the attributes of information affecting evacuation route choice and demonstrated how to use information to influence evacuees' route choice to reach the SO state (6). Furthermore, disasterand evacuation-related information available through various information dissemination channels to both evacuees and nonevacuees can affect their decisions before and during the evacuation. For example, information about the blockage or closure of a certain area would result in the diversion of nonevacuees who originally intend to traverse the (now-closed) hot zone. Nonevacuees are also likely to synthesize the closure information and past experience to determine their own best alternative route. Such critical aspects of information have not been well addressed in the literature for no-notice evacuation scenarios.The main technical contributions of this study are twofold. First, a simulation-optimization framework is proposed in which a universal quickest flow (UQF)-also known as the earliest arrival flow
Modeling of Evacuation and Background Traffic for Optimal Zone-Based Vehicle Evacuatio...