Optimal Signal Timing For Urban Intersection Based On Vehicle Emission

With the rapid increase of urban vehicle, the environmental problems caused by vehicle emission have become more and more serious, Intersection is a bottleneck of urban traffic, where vehicle’s delay, repeated acceleration and deceleration lead to large traffic emission. The study of signal control with low emissions has become one of topic research fields in intelligent transportation. Traditional emission signal control scheme integrates vehicles’stop rate and average delay, and acquire the relationship between traffic emission and signal timing in term of Webster formula. However, this approach fails to consider the initial queue length in intersection, and the Webster formula is not suitable for the high saturation degree condition. In order to overcome these drawbacks, this dissertation considers the initial queue length in intersection and random vehicle arrival, and proposes an adaptive control scheme for low emission and verifies the effectiveness by simulation experiments.The main work of this dissertation is as follows:1. The vehicle emission characteristics in different working conditions are discussed, and the vehicle emission models are utilized into developing traffic simulation software USTCMTS2.0, which provides a platform for evaluating the influence of different signal timing plans on vehicle emission.2. Q-learning is used to search for a low emission query table of optimal timing for different queue length based on traffic emission simulation. In allusion to the low learning efficiency of Q-learning, the fuzzy reasoning is introduced to accelerate convergence of Q-learning.3. An urban adaptive traffic control system is developed. The data communication module, database design, operation interface, display interface and simulation module are also described. According to queue length in intersection, traffic emission control can be realized.