基于模糊推理的驾驶员反应时间修正研究

发布时间：2018-05-30 11:08

本文选题：智能交通 + 主动防碰撞　；参考：《广西大学学报(自然科学版)》2014年06期

【摘要】：为进一步完善汽车主动防碰撞预警系统,建立了一种基于模糊推理的驾驶员反应时间修正方法。该方法以驾驶员的实时反应时间为研究对象,通过实时采集驾驶员的加油频率、深加油比例、制动频率、深制动比例等数据,利用所建立的模糊推理规则判断出一个适应于当前驾驶员动态驾驶倾向的实时反应时间,然后对Berkeley算法进行修正,计算出车辆行驶所需要的安全车距,并在驾驶模拟器上进行了对比分析。经模拟实验验证,本方法所确定的15名驾驶员的理论平均反应时间为1.000 0 s,方差为0.041 7,而实际平均反应时间为1.007 7 s,方差为0.046 9,并且根据理论反应时间所确定的预警距离与根据实际反应时间所确定的预警距离比较接近。可见本方法能在一定程度上提高安全预警算法的准确性,能够充分考虑驾驶员的实时驾驶特性,降低预警系统的虚警率。
[Abstract]:In order to further improve the vehicle active anti-collision warning system, a driver reaction time correction method based on fuzzy reasoning is established. This method takes the driver's real time reaction time as the research object, and collects the data of the driver's refueling frequency, deep refueling ratio, braking frequency, deep braking ratio and so on in real time. Using the established fuzzy inference rules, a real-time reaction time suitable for the driver's dynamic driving tendency is determined, and then the Berkeley algorithm is modified to calculate the safe vehicle distance required for vehicle driving. A comparative analysis is made on the driving simulator. The simulation results show that, The theoretical average reaction time of 15 drivers determined by this method is 1.000 s, the variance is 0.041 7, while the actual average reaction time is 1.007 7 s and the variance is 0.046 9. The warning distance determined by international reaction time is close. It can be seen that this method can improve the accuracy of the security early warning algorithm to a certain extent, fully consider the driver's real-time driving characteristics, and reduce the false alarm rate of the early warning system.
【作者单位】：山东理工大学交通与车辆工程学院;
【基金】：国家自然科学基金资助项目(61074140) 山东省自然科学基金资助项目(ZR2011EEM034)
【分类号】：U463.6;U491.25

【参考文献】