Integrated travel demand models for evacuations: a bridge between social science and engineering

Abstract

Since 9/11, the Indian Ocean tsunami and hurricane Katrina, the number of papers that are being published related to mobility simulation in evacuation conditions has signifi cantly increased. Though several topics have been developed, they tend to be implemented with an isolated and non-system approach and for specifi c kinds of dangerous events. This work aims to present a classifi cation and specifi cation of demand models for mobility simulation in evacuation conditions under different evacuation scenarios, in respect to different temporal conditions. A general framework is proposed to support the analysis of dangerous events, in respect of type and effects, especially in time. Three different temporal evolutions are identifi ed and systematized: event developments and the relative conditioning on the system; user modifi cation of behavior; and planning and management evolution. Leaving from the integrated temporal evolutions, the user behavior in the system context is analyzed and specifi c models are developed. The importance of SP surveys to analyze user behavior in evacuation conditions is highlighted and a hybrid class of surveys, termed SP with a physical check, is introduced. An integrated demand model is specifi ed and calibrated for a dangerous event with effects on travel demand, with diffuse effects in space and delayed in time, according a behavioral approach.

Transport models and intelligent transportation system to support urban evacuation planning process

“…The state of the art on models and applications concerning travel demandtransport supply interaction models in emergency conditions is presented in [8,21]. A general travel demand model for emergency conditions is proposed in [23]. Path design during an evacuation can be considered if the vehicle travels from a stop to another stop or if the vehicle must perform a route with multiple stops [24].…”

“…A dynamic path choice model is adopted to simulate supply-demand interaction. Starting from these considerations, an integrated system of models is proposed, considering an anthropic dangerous event, which mainly produces effects on demand, with diffuse effects in space and delayed effects in time [23]. The integrated system of demand models is subdivided into the following sub-models: † generation and departure time model gives the level of demand in the study area according to the reference period and user category; † modal choice/distribution model estimates the number of users undertaking a transport mode from an origin to a refuge area, form available alternatives (walking, cars, buses and emergency vehicles for particular categories of users); the distribution model simulates user distribution among different safe points, refuge areas, destinations and previously defined or not by a public decision-maker; and † path choice model estimates the probability of choosing a path and the path flow given the departure time, the origin, the destination and the mode.…”

“…The first activity concerned the calibration of a system of models to simulate the evacuation observed in the two exercises and to simulate evacuation conditions in a designed scenario. Table 1 shows the variables of the supply and demand sub-models for which parameters have been calibrated (see details in [8,23]). This activity is necessary for developing models close to the reality for support transport system design (second activity).…”

Hurricane evacuation demand models with a focus on use for prediction in future events

“…The literature on evacuation behavior is extensive and many useful reviews are available, such as Dash and Gladwin (2007), Yazici and Ozbay (2008) and Murray-Tuite and Wolshon (2013). It is also worth noting that Russo and Chilà (2014) creates a generic framework for demand modeling in evacuation situations that bridge many types of events including hurricanes. We focus in this paper on studies that relate to quantitative models of evacuation behavior, and specifically consider their suitability for use in prediction.…”

Urban exposure: training activities and risk reduction

“…By means of evacuation, the societal risk is reduced. The quantity of reduction can be estimated in two ways adopting:  modeling methods that consist on simulation of transport system in evacuation conditions adopting a set of models that comprehends: o transport demand models to simulate mobility choices [22]; o supply models to simulate available transport infrastructure and services during an evacuation [23]; o transport supply-demand interaction models in relation to route choices and then the estimation of urban network flows [24];  monitoring methods that consist of implementation of an evacuation that is monitored to acquire quantitative information about exposure quantity reduction. It is possible, in a theoretical way, to estimate, for each training action, two different risk levels:  a before quantity risk level that correspond to an emergency scenario, defined in terms of event occurrence and vulnerability, for which the evacuation training action considered has not been implemented;  an after quantity of risk level that correspond to an emergency scenario in which the evacuation training action has been implemented.…”