首页> 外文会议>International forum on automotive lighting >Heterogeneous data fusion to control a predictive bending light
【24h】

Heterogeneous data fusion to control a predictive bending light

机译:异构数据融合控制预测弯曲光

获取原文

摘要

About 90 percent of the information which is necessary in order to ensure a safe handling of a vehicle is based on visual effects.[ECK93] assumes that in about 50 percent of all traffic accidents the optical recognition plays a significant role.A variety of different authors shows,that driving through bends during nighttime presents the highest risk potential.This statement is confirmed by [SCH03] as well,according to whom the highest potential risk for an accident is to find on rural roads especially in bends.The rate of bends on German rural roads equals about 32.7 percent.Based on the above mentioned facts a predictive bending light provides a further improvement of visibility conditions in about one third of all situations which may occur during a night drive.The main objective of the predictive bending light is to improve the reduced visibility just before entering a bend.However,a problem with swiveling headlamps exists in so called S-shaped-bends.Right before reaching the inflection point the light is-due to the nature of the system-swiveled into the opposite direction,what reduces the visibility distance even more.The reason for this behavior is to find in the car's disability to predict the course of the road ahead,a bend cannot be detected prior to passing through it.Today's AFS-Systems operate based on the actual driving conditions such as the current steering wheel angle but not based on a trajectory which the car may pass through.The solution for this problem is to find in the application of anticipatory sensors enabling the predictive control of the headlamps.In this case,navigation and camera (lane-information) data are used in order to calcu-late the car's most probable path.An additional advantage is realized regarding the driver's line of sight.According to [Ewe02] a headlamp system helps even more,when the driver's line of sight behavior at nighttime equals the behavior at daytime.This can be realized by applying a predictive lighting control.In order to drive a vehicle even safer through the night it is necessary to provide an optimal visibility at all situations.
机译:为了确保安全处理车辆的安全处理是基于视觉效果的约90%。[ECK93]假设在光学识别中大约50%的流量事故中的大约50%起着重要作用。作者所示,在夜间驾驶弯曲呈现出最高的风险潜力。这句话也被[Sch03]确认,事故的最高潜在风险是在农村道路上找到尤其是弯道。弯曲率在德国农村道路上等于32.7%。基于上述事实,预测弯曲光在夜间驱动期间可能发生的所有情况中可能发生的所有情况中的大约三分之一的能见度条件进一步提高。预测弯光的主要目的是为了提高在进入弯道之前的可见性降低。然而,在达到盈盈之前,可以在所谓的S形弯曲中存在旋转前照灯的问题光点光线是 - 由于系统的性质 - 旋转到相反的方向,更加降低可见度距离。这种行为的原因是在汽车的残疾中找到预测前方道路的过程,a在通过它之前无法检测到弯道.Today的AFS系统基于实际的驾驶条件,例如当前方向盘角度,但不是基于汽车可能通过的轨迹。此问题的解决方案是找到预期传感器的应用启用了前照灯的预测控制。在这种情况下,使用导航和相机(车道信息)数据来计算汽车最可能的道路最可能的道路。在驾驶员的行中实现了额外的优势视觉。根据[eWe02],当夜间驾驶员的视线行为在白天时,前照灯系统有助于更多。这可以通过应用预测照明控制来实现这一问题。沿着驾驶车辆甚至更安全地驾驶车辆,在所有情况下都需要提供最佳的可见性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号