Study On The Optimization Methods Of Signal Timing Parameters Of Urban Traffic Control

With the sustaining and rapid development of Chinese economy, the acceleration of urbanization, and the improvement of motorization, the urban traffic problem has been the bottleneck to the sustainable development of economy and society to some extent. The Advanced Urban Traffic Control System is not only one of the important approaches to improve the operating efficiency in urban traffic, but also an important symbol of urban modernization. In China, its function is far from be sufficiently exerted. This dissertation chooses the key problem in urban traffic signal control—signal timing parameters optimization—as its research target. The study puts emphasis on the objectives, models and methods of optimization. The research findings of the dissertation contribute to a further study into the urban traffic control system, which fits the real traffic situation in China. It also serves as academic support for implementing advanced urban traffic control. It maintains high value and guidance function in scientific research. This dissertation is consisted of the following chapters interrelated to each other:(1) The basic theory of urban traffic control. This chapter first brings in the preliminary definitions used in traffic signal control, including signal phase, signal stage, cycle length, split, offset, etc. And the advantages and disadvantages of setting signal lights in intersections are discussed. The principles and methods of signal phase and phase sequence design are also involved. At last, the primary performance indexes, the secondary performance indexes, and the relationship between performance indexes and signal timing parameters are discussed. The above-mentioned points are the basis of the dissertation, as the basic theory for signal timing parameter optimization.(2)The optimization methods of cycle length. Cycle length is directly related to capacity and vehicle average delays at intersections. It is a very important parameter determining the traffic signal control benefit in pre-timed or adaptive traffic signal system. In this chapter, the general optimization methods of cycle length are discussed in detail, including TRRL method, ARRB method, HCM method and conflict point method. The merits and defects of these methods are analyzed by comparison. The traffic intension model is established in the light of two traffic flow parameters: traffic volume and time occupancy. Therefore, traffic load at the intersection can be reflected by the calibration of these two traffic flow parameters as the gist of cycle length optimization. According to the formula of time occupancy, it can be converted into the function of speed. So, the variation regularity of traffic intension obeying traffic volume is reflected by the relation between traffic volume and speed by the Edie model in traffic flow theory. The cycle length optimization model is proposed for intersections with 2～5 critical phases based on traffic intension. This method can uniquely confirm the state of traffic flow, and remedy the defects in conventional methods. It is validated by photograph investigation in three intersections in Chanchun City. The cycle length value fits the real traffic situation. The delays are low when traffic load is light, while the capacity is high in rush hours, with a four critical phases timing case presented in the demonstration. In the end, a study is made on the cycle length optimization in coordinated control, including the cycle length and double-cycle intersection. (3)The optimization methods of split. Split reflects the length of green time of each phase at intersection. The optimization of split is to reasonably distribute effective green time to various phases. In this chapter, a retrospect is first made on "equivalent saturation" put forward by Webster and "non-equivalent saturation" by Akcelik. Then a further study is on the confirming method of the critical phases. At last, the split optimization methods about overlapping phase and non-overlapping phase, and the calculation of actual green time ar...