摘要
研究驾驶员的制动驾驶行为是汽车辅助驾驶系统应用的基础。建立BP神经网络,基于QJ-4B1型6自由度动感型模拟驾驶仪平台,邀请10名志愿者模拟在城市道路上驾驶过程中车辆相对距离、前车加速度、辆车相对速度和碰撞时间倒数等制动行为特征参数,并作为BP神经网络模型的输入参数,对驾驶员的制动驾驶行为进行预测,由结果显示,在240组检测样本中只有3组数据误差绝对值超过1。由此可以看出,基于BP神经网络的驾驶员制动行为模型对于驾驶员的制动行为预测较为准确,模型有效,可以在辅助驾驶系统中得到广泛应用。
The study of driver’s braking behavior is the basis of the application of the vehicle-assisted driving system.Based on the QJ-4B1 6-DOF dynamic simulator platform,10 volunteers were invited to simulate the relative distance,vehicle acceleration,vehicle relative speed and collision time in the process of driving on the city road.And parameters are used to predict the driving behavior of the driver basing on the BP neural network model.The result shows that only three sets of data in the 240 groups have error absolute value of the error exceeding 1 it can be seen that the driver braking behavior model basing on BP neural network is more accurate and the model is effective and can be widely used in the auxiliary driving system.
作者
刘志强
王玲
贾海江
倪捷
LIU Zhi-qiang;WANG Ling;JIA Hai-jiang;NI Jie(School of Automobile and Traffic Engineering,Jiangsu University,Jiangsu Zhenjiang 212013,China)
出处
《机械设计与制造》
北大核心
2019年第6期37-41,共5页
Machinery Design & Manufacture
基金
道路交通安全公安部重点实验室(2016ZDSYSKFKT09)
