Models For the Evacuation Analysis of An Urban Road Transportation System in Emergency Conditions
A Vitetta, F Russo, University Mediterranea of Reggio Calabria, IT
Emergency conditions in transportation system can be activated by event exogenous to the transportation system (flood, electric power station, earthquake, ...) or an event endogenous to the transportation system (dangerous goods on truck, ...). The dangerous effect on the population could be immediate (for example dangerous goods on truck, earthquake) or delayed (for example flood, electric power station).
Methodologies for transportation planning in urban area when endogenous and/or exogenous events occur with delayed effect on the population have not received great attention in the literature. Only emergencies when a nuclear event occurs are treated and in some case for building and urban system. In general, there is no systematic analysis of different events connected with transportation system design. Models and algorithms specified and calibrated in ordinary conditions cannot be directly applied in emergency conditions. Users move on the network in different behavioural conditions, they do not know system congestion and reliability in real time, and the user optimum, in general, is different from the system optimum. Different models and algorithms have to be proposed in this scenario.
A problem which entails particular attention in terms of coordination and management is that strictly linked to general mobilization which is created in a densely populated area after a disaster (earthquake, flood, etc.) or in any kind of emergency. In real condition very often for verify the quality of an evacuation strategy a real simulation on the real system are experimented. This approach is very near to real condition but is expansive in term of money, organization and people to be involved. During real experimentation if some of the scenario configuration have to be modified, a new real evacuation have to be organized. The approach proposed consider a simulation approach with quantitative models where the real condition are reproduced on a computer. The computer simulation is cheaper than real simulation and different strategies and scenarios should be tested.
In this paper a general formulation for defining the risk and the relative frequency and magnitude (in term of vulnerability and exposition) is proposed. The general problem is discussed for define the measures for exposition reduction (evacuation) in a transportation system when an exogenous event happens. A general model and relative solution algorithms are proposed for design and simulate road urban network system in emergency conditions. Models and algorithms used for simulating system in ordinary conditions, updated for emergency conditions are proposed and compared. Two different approaches are considered to simulate planning co-ordination: 'what if' models (system or user optimum with deterministic user behaviour) and 'what to' models (optimal user distribution and departure time, optimal network design system). An application was developed in a transportation system considering the application of stationary and non stationary models. Pros and cons for the different models in different simulation conditions are discussed.
Association for European Transport