Abstract:

A system partition model for urban arterials is proposed based on classical MAXBAND model. The model divides the whole arterial into several subsystems, each with three to six signalized intersections. Each subsystem is optimized individually to achieve maximum two-way bandwidth. The through traffic in different subsystems is provided average green bandwidth as equal as possible. Genetic Algorithm is applied to solve the model, the optimal results implies that intersections with critical cycle length or large spacing is the partitioning point. Compared to the plan generated by Synchro, the one produced by the proposed model increases the average progression efficiency by 27.8%. CORSIM simulation is conducted to compare the performance of signal timing plans generated by the proposed model and Synchro respectively, and results analysis indicates that compared to the plan generated by latter one, the one produces by former one resulted in better performance indexes, such as average delay, average speed and stops.

Key words: traffic engineering, system partition, super-long arterial, coordination control, genetic algorithm

摘要：

针对以直行车流为主要流向，实行两相位控制的城市干线，基于经典的MAXBAND 模型，建立了干线分子区协调控制模型.该模型自动将干线分为若干具有3–6个交叉口的控制子区，每个子区追求最大化绿波带宽，不同子区协调方向的直行车辆享有尽可能均等的绿波通行时间.采用遗传算法求解模型，求解结果显示，分段点一般为周期相对较大或间距较大的交叉口.通过与Synchro 优化方案相比，本文模型优化的方案可以使子区的平均带宽有效率提高27.8%.利用CORSIM对分别采用本文模型和Synchro 划分的 6 种配时方案进行仿真，结果表明，与后者优化的方案相比，本文模型优化的方案具有更优的干线平均延误、平均车速、停车率等性能指标.

关键词: 交通工程, 子区划分, 超长干线, 协调控制, 遗传算法