当前位置:主页 > 科技论文 > 路桥论文 >

连续驾驶时间下高原驾驶员适宜性模糊聚类评价

发布时间:2018-04-16 15:09

  本文选题:高原驾驶员 + 综合评价 ; 参考:《科学技术与工程》2017年10期


【摘要】:为了使高原驾驶员适宜性评价更加客观准确,从驾驶员疲劳角度出发,将影响驾驶员疲劳的主要因素连续驾驶时间作为参照依据,运用模糊理论,对驾驶员视觉特性、速度估计特性、反应特性、处置判断特性4个方面进行综合评价。通过实地行车试验获取20名驾驶员适宜性指标数据,对其中13名驾驶员样本数据进行模糊聚类,建立模糊相似矩阵与等价矩阵,依据聚类样本驾驶员连续驾驶时间长短确定最佳阈值,并将驾驶员适宜性分为4类,最后将已聚类驾驶员样本数据作为识别标准,利用择近原则对其余7名驾驶员进行识别与评价。结果表明:将连续驾驶时间作为参照依据,利用模糊聚类理论对驾驶员适宜性进行评价的方法是可行的。
[Abstract]:In order to make the evaluation of driver suitability more objective and accurate, from the point of view of driver fatigue, the main factor affecting driver fatigue, continuous driving time, is taken as the reference basis and the fuzzy theory is used to analyze the driver's visual characteristics.The velocity estimation characteristic, the reaction characteristic and the disposition judgment characteristic are evaluated synthetically.In this paper, 20 drivers' suitability index data were obtained through field driving test, and 13 drivers' sample data were clustered by fuzzy clustering, and the fuzzy similarity matrix and equivalent matrix were established.The optimal threshold is determined according to the continuous driving time of the clustered sample driver, and the driver suitability is divided into 4 categories. Finally, the cluster driver sample data is used as the identification standard.The other 7 drivers were identified and evaluated by the principle of proximity selection.The results show that it is feasible to use fuzzy clustering theory to evaluate driver suitability based on continuous driving time.
【作者单位】: 新疆农业大学机械交通学院;
【基金】:国家自然科学基金(51168045)资助
【分类号】:U491.25

【相似文献】

相关期刊论文 前5条

1 惠先宝;高速公路管理机构设置方法及其应用[J];交通标准化;2000年03期

2 姚磊;刘渊;;基于改进模糊聚类与ANFIS的高速公路事件检测[J];计算机工程与应用;2013年19期

3 张恒才;陆锋;陈洁;;微博客蕴含交通信息的提取[J];中国图象图形学报;2013年01期

4 郭洪洋;张玺;刘澜;刘海旭;闫锴;;多源信息约束下的路径选择模型与算法[J];计算机应用;2014年07期

5 ;[J];;年期



本文编号:1759446

资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/daoluqiaoliang/1759446.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户10c6d***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com