| 摘 要: 在汽车驾驶行为分析中，一般通过车载传感器收集数据，并对数据进行处理，但是车载传感器收集的数 据具有很多弊端，比如具有数据量大、无法分辨方向等。本文提出一种基于基准点映射的驾驶行为分析算法：根据原始 行驶记录，使用一定规则和Geohash算法建立带有方向的基准点，并将行驶记录投影到对应的基准点中，采用k-Means 聚类算法对基准点数据进行分类，减少有效数据点个数，为后续危险驾驶行为分析提供基础。 |
| 关键词: 基准点；驾驶行为；Geohash算法；k-Means聚类 |
|
中图分类号: TP399
文献标识码: A
|
| 基金项目: 教育部人文社会科学研究青年基金项目(15YJCZH005)；辽宁省自然科学基金项目(20180550018)；2018年度辽宁省普通高等教育本科教学改革研究项目(544)；2017年 大连外国语大学科研基金项目成果(2017XJYB17). |
|
| An Analysis of Driving Behavior Based on Reference Point Mapping |
|
GUO Qiaoli,CAO Hongju1,2
|
1.( 1.Dalian Neusoft University of Information, Dalian 116023, China;2.
2.Dalian University of Foreign Languages, Dalian 116044, China)
|
| Abstract: In the car driving behavior analysis,data collection and data processing are generally performed by means of vehicle sensors.However,there are many drawbacks for the collected data by vehicle sensors,such as a large amount of data and indistinguishable directions.This paper proposes a driving behavior analysis algorithm based on reference point mapping :According to the original driving record,a certain rule is used to establish the reference point with direction.The driving record is projected into the corresponding reference point classified by k-Means clustering to reduce the numbers of effective points,which is prepared for the subsequent cluster analysis and the dangerous driving behavior analysis. |
| Keywords: reference point;driving behavior;Geohash;k-Means clustering |
