基于多分辨率建模的车路协同系统仿真场景设计与实现

MR-based CVIS Scenario Design and Implementation

车路协同技术是当今国际智能交通领域的前沿技术,是城市交通问题的有效解决方案。多分辨率建模方法是研究复杂系统仿真的一种重要手段,特别适用于车路协同系统的复杂性、层次性特性。基于多分辨率思想,从交通流信息、多车与地面设备间的信息交互、单车的车车、车路信息交互三种层次构建了车路协同系统多分辨率模型以及基于高层架构体系(HLA)的车路协同系统仿真框架;通过分析车路协同系统多分辨率层次需求,从高分辨率及中分辨率下的车队排队时间延误、低分辨率下路段平均行驶时间建立完整的路段关联度体系,可以从多层次、不同分辨率对车路协同系统中路段间关联度进行更加准确的描述;结合遗传算法将车路协同系统解聚为动态控制区域,分析了不同分辨率下的子系统和不同模式的数据交互信息及其耦合关系,实现区域最佳协调控制效果。从多层次、不同分辨率更加准确地定义了路段关联度,更好地进行动态控制子区的划分。

Cooperation Vehicle-Infrastructure System （CVIS） is the cutting-edge technology in current international intelligent transportation field and effective solution for urban transportation problems Multi-Resolution Modeling （MRM） is one important means for complex system simulation research, especially fits for the complexity and hierarchical characteristics of CVIS. Models of traffic flow information, information interaction between multi-vehicle and ground equipment, vehicle-vehicle, vehicle-road information interaction, and High Level Architecture （HLA） -based CVIS simulation framework were built up. Through analysis of vehicle line-up delay time, average driving time in low resolution, a multi-resolution correlation system was established; a more specific description of correlation among CVIS links was given in multi-level and multi-resolution. Combined with Genetic Algorithm （GA）, CVIS could be disaggregated into dynamic control areas. The results show an accurate definition of links correlation, a better dynamic control subareas division, an effective reduction of average delay time in the coordinated area, and realization of optimized coordinated control.