一种虚拟交通环境中的微观交通仿真模型

为提高虚拟交通环境中车辆行为仿真的逼真度及解决用户与系统的实时交互问题,提出了一种微观交通模型.该模型由道路模型和车辆行为模型构成:道路模型包括了车道的交通属性、几何形状、路网拓扑关系等数据;车辆行为模型考虑驾驶员的交通性格、车辆性能等因素,包括自由行驶、跟驰及换道模型.车辆行为模型可检测与用户所驾驶的汽车仿真器的位置关系,并作出减速、换道、停车等响应.建立了包含153个车道的虚拟交通环境,对模型进行测试,结果表明:所建立的微观交通仿真模型能够反映仿真车辆的个体行为差异;路网中仿真车辆为600辆时,仿真周期约2~4 ms,能实现实时仿真及与用户的实时交互. 

Microscopic Traffic Model in Virtual Environment

In order to improve the simulation fidelity of vehicle behaviors in a virtual traffic environment and realize real-time interaction between users and the system, a microscopic traffic model was proposed. This model is composed of a road model and a vehicle behavior model. The road model has a hierarchy structure and includes traffic properties, geometrical parameters and network topology relationship of lanes. The vehicle behavior model takes into account traffic characters of drivers, vehicle performances and other factors, including free driving, car-following and lane-changing models. Simulation vehicles detect position relations with the car simulator driven by a user and then select to brake, change lanes or stop. A virtual traffic environment including 153 lanes was established to test the proposed model. The experiment result indicates that the microscopic traffic model can show individual differences of vehicles' behaviors. When 600 simulation vehicles are included in the road network, the time-consuming of one simulation loop is 2 to 4 ms to meet the requirements of user real-time interaction.