中文摘要：

针对传统区域交通控制技术无法应对机非冲突干扰的问题,结合中国城市道路混合交通流的特点,研究了交叉口与路段非机动车对机动车的干扰。分析了区域路网机动车交通特征,确定了混合交通特性相似的区域。基于路段非机动车的阻滞作用,分析了交叉口通行能力的折减与相邻交叉口相位差的优化。以区域路网机动车总延误为优化目标,建立了非机动车影响条件下的区域交通信号控制优化模型,优化了信号周期时长、绿信比和相位差等参数,并利用遗传算法求解模型。利用VISSIM仿真软件,以上海市杨浦区五角场环形区域路网为例对优化模型进行验证。验证结果表明：现状信号控制方案下区域路网7个交叉口机动车的车均延误为24.5-42.9s,平均为35.99s,路网总延误为256.39h,优化后交叉口的车均延误为21.8-36.4s,平均为30.12s,路网总延误为214.57h,7个交叉口车均延误减少了10%-24%,平均为16.31%。可见,优化模型能够显著降低区域路网车均延误与总延误,提高区域路网通行效率。

英文摘要：

Considering that traditional regional traffic control technology could not cope with the interference problems of motor vehicles and non-motorized vehicles conflict, the characteristics of mixed traffic flow in China＇s urban road were combined, and the interferences of non-motorized vehicles on motor vehicles at intersections and links were researched. The traffic characteristics of motor vehicles in regional road network were obtained, and the regions with similar mixed traffic characteristics were defined. Based on the blockade effect of non-motorized vehicle, the capacity reduction of intersections and phase difference optimization of adjacent intersections were analyzed. The total delay of motor vehicles in regional road network was used as optimization objective, the optimal model of regional traffic signal control influenced by non-motorized vehicles was built, and signal cycle length, split ratio and phase difference were optimized. Genetic algorithm was introduced to solve the optimal model. Taking the Wujiaochang annular regional road network in Yangpu District of Shanghai as an example, the optimal model was verified by using VISSIM software. Verification result indicates that for signal control scheme beforeoptimization, the average delays of each motor vehicles are 24.5-42.9 s at 7 intersections, the average value is 35.99 s, and total vehicle delay is 256.39 h in regional road network. For signal control scheme after optimization, the values are 21.8-36.4 s, 30.12 s and 214.57 h respectively. The decrement rate of average each vehicle delay is 10%-24% at 7 intersections, the average value is 16.31%. So the optimal model can effectively reduce each vehicle delay and total vehicle delay in regional road network, and the operational efficiency increases in the regional road network significantly. 4 tabs, 4 figs, 26 refs.