驾驶员纵向制动模型的研究

Study on the Vehicle Driver Model of Longitudinal Braking Model

基于在中国5个城市开展的自然驾驶试验,利用视觉传感器以及Matlab图像处理方法获得驾驶员在真实危险工况下的车速控制行为数据,基于驾驶员及车辆数据对驾驶员危险感知、反应时间、制动感知等行为进行分析及建模。最终利用动态神经网络建立了驾驶员纵向制动模型,为速度控制类ADAS系统的测试提供驾驶员方面的参考,并为此类系统能更好地适应中国道路交通环境提供驾驶员模型支持。

Based on the natural driving tests conducted in five cities in China,visual sensor and Matlab image processing methods are used to obtain the driver's speed control behavior data under real dangerous conditions.Based on the driver and vehicle data,the driver's perception of danger,reaction time,analysis and modeling of braking perception and other behaviors.Finally,a dynamic neural network is used to establish a driver longitudinal braking model,which provides a driver's reference for the testing of speed-controlled ADAS systems and can be adapted to such systems.China's road traffic environment provides driver model support.