基于公交运营异常事件快速定位存储平台设计  

Design of Storage Platform for Fast Locating Exceptional Events Based on Running Buses

作　　者：迟剑 刘艳飞 Chi Jian;Liu Yanfei(School of Mathematics and Computer Science,Hebei Normal University for Nationalities,Chengde 067000)

机构地区：[1]河北民族师范学院数学与计算机科学学院,承德067000

出　　处：《现代计算机》2022年第24期110-115,共6页Modern Computer

基　　金：河北民族师范学院2018年度校级基金项目(PT2018010);河北省省级科技计划资助项目(20310301D)。

摘　　要：公交运营中异常事件频发影响出行用户舒适度,增加公交公司运营成本。快速处理运营异常事件是目前公交公司亟待解决的问题之一。异常事件发生的不可控性以及不确定性是研究中监测难点。在完整追踪某一线路所有运营车辆的营运过程,提出了对异常事件影响指标分级,设计二级位图索引,快速定位异常事件发生车辆和路段的问题解决方法。Hadoop分布式存储框架中,车辆信息和查询地点作为Key值,设计辅助多维位图索引,构建Hbase异常数据存储库,依据Impala快速定位查询点。对比应用MapReduce计算框架和直接Impala查询完全数据集,异常特征数据快速定位性能提升明显,也减少了Impala查询存在的一些缺陷问题。Frequent abnormal events in transit operation affect passengers comfort level and increase the operating cost of bus companies. At present, the transit operation urgently needs to handle operational anomalies. The uncontrollable and uncertain occurrence of abnormal events are the difficulties of monitoring in research. Complete tracking of the operation process of all vehicles operating on a certain line, a method which grade the impact index of abnormal events and design the secondary bitmap index,quickly locate the vehicle and road segment where abnormal events occurred. In Hadoop distributed storage, vehicle information and query location are used key values, a Hbase exception data repository with multidimensional bitmap-index is builded. Compared with MapReduce and querying the complete dataset with Imala, the performance of fast location of exception data is improved apparently and reduce vulnerabilities of impala.

关 键 词：公交数据云存储平台 异常事件 Impala平台 位图索引 轨迹数据 

分 类 号：U491.17[交通运输工程—交通运输规划与管理] TP311.13[交通运输工程—道路与铁道工程]