Pick-UP Location And Bus- Routing Model For Bus-Based Emergency Evacuation

Emergence events by manmade or nature-made have been occurred frequently with the conflicting from nature and society. We need to organize some large-scale regional evacuation and take dangerous people to safe area when disaster happens, in order to ensure the safety of personnel. During the evacuation, part of people could leave by vehicles which would increase the demand of traffic network in short-term. However, in China, the high proportion of low-mobility people which have serious vulnerability do not have reliable traffic media to leave. It would help the increase of efficiency if they take public transit to evacuate. This paper built mathematical model and designed relevant solution algorithm, focuing on the problems of pick-up location and bus routing for bus-based evacuation planning, which may provide support for large-scale regional evacuation by bus.Firstly, the paper analyzed the influencing factors of pick-up location problem, including characters of event and evacuation area, and the information of event, evacuation area and evacuation demand which has to obtain before pick-up location selecting. Then it is assumed that the alternative pick-up locations with capacity and position already known would not be destroyed by disaster, and people in each evacuation cell would get to the only one corresponding pick-up location by shortest path. On the basis of these hypothesis, a bi-objective optimal model with prime objective of shortest total walking distance and secondary objective of least arterial road crossing times subjecting to capacity limited and longest walking distance was built to search for the appropriate pick-up locations. The genetic algorithm was applied to find superior scheme in Pareto optimal solutions.Secondly, after analyzing normal traffic assignment model, we regarded the distribution of buses on network during evacuation as system optimum status. Considering this, for the law of people arriving is changing by time, evacuation time was separated into several equal intervals and a linear bus routing model was built which aimed to minimize the bus traveling distance. Use primal dual interior point method to solve the model, which decrease iterations and increase the efficiency.Finally, we take Beilun District, Ningbo City as an example. It is assumed that people in the district need to be evacuated before hurricane comes. We use our model to solve the problem, also verify the model. Then we make an evacuation plan and traffic control treasure based on model results. The results indicated that the model could provide theoretic evidence and technique support for bus-based evacuation plan.