交通运输系统工程与信息 ›› 2017, Vol. 17 ›› Issue (1): 82-90.

• 智能交通系统与信息技术 • 上一篇    下一篇

三车道动态环境下汽车驾驶倾向性的 转移概率及其预测

张敬磊1 ,王晓原*1, 2,王梦莎1,王云云1,刘亚奇1,尹超1   

  1. 1. 山东理工大学交通与车辆工程学院,山东淄博255091;2. 清华大学汽车安全与节能国家重点实验室,北京100084
  • 收稿日期:2016-05-10 修回日期:2016-08-23 出版日期:2017-02-25 发布日期:2017-02-27
  • 作者简介:张敬磊(1979-),男,山东高密人,副教授.
  • 基金资助:

    国家自然科学基金/National Natural Science Foundation of China (61573009);山东省自然科学基金/ Natural Science Foundation of Shandong Province (ZR2014FM027);山东省高等学校科技计划/ Project of Shandong Province Higher Educational Science and Technology Program (J15LB07);汽车安全与节能国家重点实验室开放基金/State Key Laboratory Open Foundation of Automotive Safety and Energy /(KF16232);山东省社会科学规划研究项目/ Social Science Planning Project of Shandong Province (14CGLJ27).

Transition Probability and Prediction of Automobile Driver's Propensity under Three-lane Dynamic Conditions

ZHANG Jing-lei1,WANG Xiao-yuan1, 2 ,WANG Meng-sha1, WANG Yun-yun1, LIU Ya-qi1, YIN Chao1   

  1. 1. School of Transportation and Vehicle Engineering, Shandong University of Technology, Zibo 255091, Shandong, China; 2. State Key Laboratory of Automotive Safety and Energy, Tsinghua University, Beijing 100084, China
  • Received:2016-05-10 Revised:2016-08-23 Online:2017-02-25 Published:2017-02-27

摘要:

驾驶倾向性是汽车行驶中驾驶员情感偏好等特征的动态测度,易随环境的改 变而变化,并影响驾驶员意识和汽车操控行为;另一方面,其类型和转移概率又同时受到 后者的作用与影响.准确揭示动态环境下驾驶倾向性转移概率,对实现汽车自动驾驶和辅 助驾驶系统具有重要意义.本文以三车道为例,从环境变化,特别是从刻画交通态势复杂 性的车辆编组关系演化角度出发,设计三车道条件下的驾驶实验,采集不同行驶环境下 驾驶员倾向性并进行统计分析,揭示环境嬗变情况下,汽车驾驶倾向性转移概率.验证结 果表明,利用倾向性转移概率对汽车驾驶员倾向性进行预测的结果与实时辨识结果相吻 合,平均准确率达80%以上.

关键词: 智能交通, 驾驶倾向性, 状态转移概率, 隐马尔科夫模型, 车辆编组关系

Abstract:

Driver’s propensity is a dynamic measurement of controller’s affection, preference and others on process of driving. It will shift with environment during driving, and also affects partly the states of driver’s consciousness and driving behaviors. On the other hand, its formation and transition probability are affected by the latter all the time. It plays a significant role for researching the active driving and auto-driving systems to reveal the transition probability exactly of vehicle group relationship in complex environment. Take three-lane condition as an example, data of driver’s propensity can be collected and analyzed through driving experiments in different environments from the angle of environmental change, especially from the angle of evolution of vehicle group relationship that reflect the complexity of traffic situation. Therefore, driver’s propensity transition probability can be revealed under the evolution of complex environment conditions. Verification results show that the predictive outcomes that gotten by transition probability are consistent with that of real-time recognition, the accuracy rate is more than 80%.

Key words: intelligent transportation, driver’s propensity, state transition probability, hidden Markov model, vehicle group relationship

中图分类号: