«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

32

期数:

2022年01期

页码:

61-66

栏目:

大数据分析与挖掘

出版日期:

2022-01-10

文章信息/Info

Title:

A Trajectory Partition Method Based on Double Velocities

文章编号:

1673-629X(2022)01-0061-06

作者:

祝 贺; 于子兴

南京邮电大学 计算机学院,江苏 南京 210023

Author(s):

ZHU He; YU Zi-xing

School of Computer Science and Technology,Nanjing University of Posts and Telecommunications,Nanjing 210023,China

关键词:

数据挖掘; 轨迹划分; 移动特征; 双速度特征; 停留点提取

Keywords:

data mining; trajectory partition; movement features; double velocities features; resident points acquisition

分类号:

TP302. 7

DOI:

10. 3969 / j. issn. 1673-629X. 2022. 01. 011

摘要:

轨迹数据挖掘对于基于位置的应用非常重要,而轨迹划分是轨迹数据挖掘的重要步骤。 节点的运动轨迹数量很大,轨迹形状迥异千差万别,使得轨迹划分成为轨迹数据挖掘的关键和难点。 轨迹划分的目的是去掉多余的轨迹点,留下重要的轨迹点数据,且要求处理后得到的轨迹留有原来轨迹的特征。 该文从速度和加速度等方面分析了节点的运动行为,提出了一种基于双速度特征的轨迹划分方法( trajectory partition method based on double velocities,TPDV)。 在 TPDV 中,首先通过检测节点移动速度的变化来找出速度改变点,并且根据节点加速度变化也可提取出特征点,然后在检测节点的速度和加速度变化的前提下,根据节点活动的时间和范围来确定停留点,最后根据提取的特征点对子轨迹进行划分。 基于 Geolife 轨迹数据集的仿真结果表明,基于双速度特征的轨迹划分方法在运行时间、简化率和划分误差方面都表现较好。

Abstract:

rajectory data mining has become an increasing concern in the location-based applications,and the trajectory partition is takenas the primary procedure of trajectory data mining. The amount of movement trajectories of nodes is typically large,and the trajectoryshapes are extremely different from each other, which makes the trajectory partition be a key problem to the trajectory data mining. Thepurpose of the trajectory partition is to retain the vitaltrajectory data while removing the redundant data in the trajectories, so that thesimplified trajectories retain the characteristics of the original trajectories. In this paper,the movement behaviors of nodes are analyzedfrom the aspects of moving speeds and moving acceleration, and then a novel trajectory partition method based on double velocities( TPDV) is proposed to partition the trajectories. In TPDV,we first extract the switch points in which the movement speeds of nodes arevaried significantly. Then we extract the change points where the accelerated speeds of nodes are varied significantly. Finally, bydetecting the variations of movement speeds and accelerated velocities of nodes,we obtain the resident points according to the time andmotion ranges. According to the extracted feature points, we can partition the sub - trajectories. Simulations on the Geolife trajectorydataset indicate that TPDV can achieve a preferable trade-off among the simplification rate,running time and error rate.

常用功能

更新日期/Last Update: 2022-01-10