基于多源信息融合模型的高龄驾驶员安全性评估方法研究

Older Driver's Behavior Safety Evaluation Methods Based on Multi-source Information Fusion Model

针对老年人驾驶中遇到的问题,本文设计了一个适合于高龄驾驶员的辅助驾驶系统。首先分析了高龄驾驶员驾驶行为随着驾驶速度和持续驾驶时间的关系。接着,结合路面信息和驾驶行为的特征,采用模糊信息融合方法给出了驾驶环境安全度的综合评价。通过模糊信息融合推理,可以使高龄驾驶员对自己的驾驶行为和周围状况有一个全面的认识。随后通过对驾驶行为和事故日志的挖掘,寻找不同驾驶员的特征,通过调整推理规则,提供适合于不同驾驶员的评价。本文所设计的辅助驾驶系统充分考虑了高龄驾驶人员的行为特征,实时为高龄驾驶者提出驾驶行为调整策略,全面提升高龄人员驾驶行为的安全性和舒适性。

In this paper we propose a information fusion model to provide exact evaluation of the driver condition and improve the driver safety degree.Firstly,the influence of age and driver speed on the driver safety is analyzed,and we found the safety degree declined as the exponential function.Secondly,based on the driver's behavior information data and the road information data,we proposed a fuzzy information fusion method to give the integrated estimation of the driver safety degree.In this way,we can get them comprehensive understand about the surroundings and themselves.Finally,the information fusion rules can be adjusted to satisfy different drivers through analysis the driver behavior log and their history incident log.Based on the fuzzy inference results,some driver behavior adjust policies can be proposed to the drivers.The experimental results based on data form American National Highway Traffic Safety Administration show that the methods proposed in this paper is more sensitive and is efficient in improving the older driver safety.