摘要
简述当前开展面向安全预警的机动车驾驶意图研究的目的和意义。分析国内外研究现状,得出从驾驶员行为及驾驶动作序列角度开展驾驶意图研究的可行性和有效性,同时介绍了2种基于概率与数理统计的机动车驾驶意图建模方法。结合驾驶员在直线封闭路段实施驾驶行为特征,阐述使用隐马尔科夫模型(HMM)理论建立驾驶意图模型的步骤以及模型参数学习和系统在线优化算法等内容。对驾驶意图模型网络结构、动态性能方面相关研究方向进行展望。
The purpose and significance of research on driver's intention for traffic safety precaution are described.Based on the analysis of related research status at home and abroad,the approach to the research on driver's intention by means of driver's behaviors and driving action sequences is feasible and valid.Two modeling methods for driver's intention are introduced based on probability and mathematical statistics theories.Considering characteristics of driver's behaviors implemented in Linear closed road,this research presents the hidden Markov model(HMM)-based procedure of driver's intention modeling as well as the model parameters learning and system online optimization.The research fields such as the network structure and dynamic property are discussed and considered as the subsequent importance with HMM-based driver's intention model.
出处
《交通信息与安全》
2012年第3期87-92,共6页
Journal of Transport Information and Safety
基金
国家自然科学基金(批准号:60974094
61104158)
教育部工程研究中心开放基金项目(批准号:WHUTERCTS2011A09)资助