基于模糊聚类分析的车道变换阶段划分  

Division of lane change based on fuzzy clustering analysis

作　　者：詹盛[1] 徐远新[1] 石涌泉[1] 王畅[1] 

机构地区：[1]长安大学汽车学院,陕西西安710064

出　　处：《计算机工程与设计》2013年第9期3293-3297,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(51178053;50678027);十一五国家科技支撑计划基金项目(2009BAG13A05)

摘　　要：车道变换过程的精确划分对车辆智能控制研究具有重要影响,细化车道变换的各个阶段能够为换道决策提供可靠理论支撑。针对车道变换的不同阶段以及车辆各个特征参数变化规律存在的差异性,基于实车试验,提出利用模糊聚类分析对车道变换过程进行划分。在分析各个换道指标的基础上,确定方向盘转角和车辆横摆角速度作为换道表征指标,考虑到方向盘转角和车辆横摆角速度的在性态和类属方面存在着中介性,具有亦此亦彼的特性,因此,选用聚类分析中的软划分对实车换道过程进行在线划分。该分类方法表现出良好的可分性。The accurate division of lane change has certain influence on vehicle intelligent control research. Refining lane change can provide reliable theoretical support for lane changing decision. The change of vehicle characteristic parameters during the dif- ferent lane change stage exists certain differences. The division of lane change process is performed by using of fuzzy cluster anal- ysis on basis of the real vehicle experiment data. After the various indexes of lane change analyzed, the steering wheel angle and yawrate of vehicle are determined to represent lane change, because there are betweenness both in performance and class attri- bute, in addition, these have the also this al^o peter, therefore, soft partitioning to divide lane change process is taken. Finally, the division of the real ear is carried out online. This classification method shows good divisibility.

关 键 词：智能控制 细化 车道变换 模糊聚类 阶段划分 中介性 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]