Simulation Models And Optimal Control For Traffic Flow Evolution On Urban Expressway System Under Accident Conditions

Urban expressway plays an important role in urban road network. Traffic accident will cause serious traffic congestion. Studying traffic flow characteristics influenced by accident and identifying the source of traffic congestion in the bottleneck caused by accident will be helpful to find out the traffic control measures and provide scientific theory basis for traffic practice. This paper models traffic flow in basic section and weaving section of urban expressway influenced by accident using cellular automata approach and presents some effective traffic control measures. Cellular automation traffic flow models, with flexible rules, good expansibility, have the advantage of modeling microscope property of traffic flow under accident conditions. The contents of this dissertation are as follows:1. According to the two-lane and three-lane cellular automaton model, a new series of merging rules and changing lane rules upstream the accident sites are added by considering the empirical driving behavior. The flux, averaged velocity, capacity and the spatial-temporal evolution features are analyzed. The results suggest that percentage of two-lane expressway with one lane blocked capacity available is35%; percentage of expressway with one lane and two lanes out of three lanes blocked capacity available is49.2%and17.1%, which is consistent with the results in HCM2000.2. Considering speed limit and the interactions of vehicles updating in appropriate section of road, cellular automaton traffic flow control model are established base on the two-lane and three-lane cellular automaton model under accident condition. By selecting appropriate values of the restricted velocity and length of speed limit region upstream the accident, the traffic stream passing through the accident sites is restricted and traffic jams is alleviated. By setting up properly the interactions of vehicles updating section upstream the accident, vehicles could smoothly passing the accident position in order. The simulation results show that this control measures could enhance the capacity of system and raise the averaged velocity when traffic is in congested state. In this way, traffic congestion upstream the accident could be eased in some extent.3. The weaving section, as a typical bottleneck, is one source of vehicle conflict and an accident-prone area. When an accident happens, some lanes will be blocked and the road capacity will be reduced. This will cause more complex traffic dynamics. A cellular automaton model to characterize accident-induced traffic behavior around the weaving section are presented. An effective control measures is proposed and verified for traffic management under accident condition. The spatial-temporal profiles, the density and velocity distribution diagrams are presented to show the impact of accident. It was shown that the proposed control measures can improve the capacity of weaving section under both normal and accident conditions; the accident occurring on lane1(median lane) in the weaving section are more inclined to cause traffic jam and reduce road capacity; the capacity of weaving section will be greatly reduced when the accident happens downstream the weaving section.