Development of In-Vehicle Collision Warning System for Two-Lane Highways Considering Driver Characteristics

碰撞预警系统（CWS）通过测量与最接近的车辆的距离（在双车道高速公路上）并估计其速度来防止碰撞，以便在安全的情况下提醒驾驶员超车。建议项目将进一步发展申请人最近在博士研究中开发的CWS。该项目分为三个阶段：(1)进一步开发车辆动力学和驾驶员仿真模块；(2)感知策略和CWS算法的发展；(3)对完整的虚拟系统样机进行仿真测试。该项目的第一阶段涉及使用基于作者铰接车辆和suv动态模型的simulink环境。为了自动控制模型中的牵引和制动特性，推进和制动子系统将被添加到原始模型中。将开发车辆处理模型，包括用于传感器障碍物检测的ADAS选项。驾驶员模型将包括转向控制、牵引/制动控制、决策和感知/反应模型。该项目的第二阶段涉及完成超车动作所需时间与预测碰撞时间之间的比较。与标准方法不同，我们建议使用车辆的技术参数而不是统计时间估计。这是由于个体设计特征和外部条件对超车指标的影响很大。集成的ADAS系统将通过雷达、摄像头和超声波传感器进行三级扫描。这将在状态流环境中执行。该项目的第三阶段涉及使用设备原型的虚拟测试。主要目标是检测车辆、道路和车道，并调整算法，以适当的公差预测必要的超车时间。

Collision Warning Systems (CWS) serve to prevent collisions by measuring the distance to the closestopposing vehicle (on a two-lane highway) and estimating its speed to advice drivers when it is safe to overtakea slow moving vehicle. The proposed project will further develop an a CWS recently developed in a PhD studyby the applicants. The project involves three stages: (1) further development of vehicle dynamics and driversimulation modules; (2) development of sensing strategy and CWS algorithm; and (3) testing of the completevirtual system prototype using simulation. The first stage of the project involves the use of the Simulinkenvironment based on the authors' dynamic models for articulated vehicles and SUVs. To automatically controlthe traction and braking properties in the model, the propulsion and brake subsystems will be added to theoriginal models. A vehicle handling model will be developed, including the ADAS options for obstacledetection with sensors. The driver model will include steering control, traction/braking control, decisionmaking, and perception/reaction models. The second stage of the project involves a comparison between thetime needed for the completion of the overtaking maneuver and the predicted time to collision. Unlike thestandard approaches, we propose the use of the vehicle's technical parameters instead of statistical timeestimates. This is proposed due to the significant influence of individual design features and externalconditions on the overtaking indices. The integrated ADAS system will use three-level scanning via radar,camera, and ultrasonic sensors. This will be executed in the State-flow environment. The third stage of theproject involves virtual tests using the device prototype. The main objective is to detect vehicles, roads andlanes and to adjust the algorithm to predict the necessary overtaking time with appropriate tolerance.