Design And Simulation For Real-time Traffic Flow Control Model Research Based On Traffic Light

Traffic congestion prevents the development of city:(1) Reducing the working time of citizens; (2) Causing environment pollution. (3) Hard dealing with urgent traffic situations. The special vehicles often encounter troubles when running emergent tasks, since the traffic light control system to date can not distinguish special vehicles from normal ones. It can not dynamicly control the traffic flow and thus timely turns green while the special vehicles reach the crossing. So the traffic light control system, as a management tool of city traffic, can adjust the duration of red and green light according to the sensed traffic flow, and thus optimize the traffic controlling, solve the traffic congestion problem and let special vehicles have green lights all the way while operating urgent tasks.To solve the problem of intersection congestion control, we apply neutral network to train the congestion approach function. Then we build a fuzzy control model. By this way, it can observe the conjunction, reschedule the durations of traffic lights by’experience’, and control the traffic flow to avoid traffic jam. We build an event-driven real-time system based on hybrid automata, which reconstruct the situation when special vehicles run into light control traffic system. To guarantee the priority of special vehicles to pass and minimize the waiting time for vehicles at the direction against those of special vehicles.In the environment of Vehicular Ad Hoc Network, in order to meet the demand of real time for light control system to monitor the special vehicles, we apply IEEE 1609.4 protocol in the system. In the scenario of multi-hop transmission, middle vehicle nodes can provide a priority forward service for the messages, which are generated by special vehicles. We use UPPAAL tool to model the system and verify relevant properties. The traffic light controller in each intersection needs to deal with a large number of messages, which are produced by neighbourhood vehicles. We propose a method, which is called PR-double queue method, to process data in two queues with different priorities, and restrict the max message number with the same priority to process at each loop. By this method, the controller can not only monitor the special vehicles in real time, but also observe the dynamic change of vehicle situation around the intersection.We develop the corresponding visual traffic simulation platform based on the theory and design proposed in this thesis. This platform can simulate several scenarios that possibly happen in reality, and can check the functionality of the control algorithm for solving practical problems. Differing from other platforms, this one supports users to set parameters and to add priority attributes of vehicles to the operational module of this platform. The operational module is developed by NetBeans tool. The platform separates the module of controling signal lights from the operational module and uses SystemC language to encode the software and hardware, by which we can provide guide and help for the subsequent development.