Abstract: It is an efficient method to manage traffic flow genarated in an organized regional evacuation from two aspects: demand regulation (means to regulate the departure mode of evacuation vehicles) and supply management (means to control traffic flow in road network). Based on the bi-level decision relationship between departure arrangement and traffic control in road network, a bi-level programming model is presented to optimize the departure plan and control parameters of intersections simultaneously. The green ratio of signalized intersections appears as the decision variable of upper level to minimize the average delay, which will influence some supply characteristics such as intersection capacity; the departure schedule and routes of evacuation vehicles appear as the decision variable of lower level to minimize the total evacuation time, which will affect some demand characteristics such as flow through intersections. A genetic-based algorithm is presented, and a numerical example is given. By comparing the optimization results of the bi-level model and the results got from the optimization mode that only consider the control parametar, it is shown that the bi-level model really holds advantage in decreasing delay and evacation time.

Key words: traffic evacuation, departure arrangement, traffic control, integrated optimization, bi-level programming