距离和属性结合的轨迹数据公共子模式发现  

Common sub-pattern discovery of trajectory data combination distance with attributes

作　　者：张文涛[1] 夏战国[1] 张磊[1] 夏士雄[1] 

机构地区：[1]中国矿业大学计算机科学与技术学院,江苏徐州221116

出　　处：《计算机工程与设计》2011年第7期2447-2450,共4页Computer Engineering and Design

基　　金：国家自然科学基金项目(50674086)

摘　　要：分析了传统公共子模式发现方式中存在的问题,提出了一种距离和轨迹属性相结合的公共子模式发现算法。该算法基于平移的最小Hausdorff距离的思想,通过计算时序点集合和参考点的属性阈值找出其中的公共子模式,从而在一定程度上减少了公共子模式的发现个数,提高了实验结果的准确度,使发现到的公共子模式更具有实际意义。并对其进行了仿真实验,实验结果表明了算法的有效性。To analyze the questions that exist in the traditional way of finding the common sub-pattern trajectory data, a common subpattern trajectory data discovery algorithm combination of distance and trajectory properties is presented. The algorithm which is based on the idea of the minimum Hausdorff distance, find the common sub-pattern by calculating the timing points collection and the threshold attribute of the reference point, which reduces the number of the finding common sub-pattem to a certain extent, enhances the experimental result accuracy, and makes that the discovery ofthe common sub-pattern has the practical significance. The simulation experiment results show that the algorithm is effective.

关 键 词：轨迹数据 公共子模式 局部匹配 HAUSDORFF距离 轨迹点属性 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]