Freeway Ramp Control Method

Urban expressways are of great importance in urban transportation system. However, with the problem of bad traffic congestion number, urban traffic efficiency is seriously reduced. Therefore, it is serious to control urban expressway. The control system of urban expressways is consisted of off-ramp control, on-ramp control, at-grade intersection control and overpass control. This thesis study off-ramp control in-depth.In this thesis, urban expressway off ramp is investigative object. This thesis study off-ramp control in-depth. First, it has determined off-ramp to signal timing program which is based on analysis of the causes to traffic congestion of off-ramp and traffic characteristics of off-ramp. Then, according to particularity of Chinese traffic environment, establishment traffic simulation software based on cellular automaton. At the same time, to prove validity of this software by studying on the actual traffic data.Currently, the main control methods of off-ramp on Beijing is timing control method. However, the identical place's traffic flow in the different time is changing, which is not meet by timing control method. To solve this problem, this thesis uses real-timing control to control off-ramp. Real-timing control on off-ramp is control method which is unit of Q-learning algorithm and extension control. The key of control algorithm is to obtain reward function of Q-learning algorithm by using extension control. Then, to determine the state of off-ramp lights according to this reward function.The simulation which is cellular automaton traffic simulation software established in this thesis indicates:When off-ramp uses real-timing signal control, the exit ramp control signal in real time, queuing vehicles on the off-ramp are less than the number of vehicles when off-ramp uses timing signal control. Therefore, real-timing signal control can ensure vehicles of main road priority to pull out, and alleviate traffic congestion of the expressway to a certain extent.