摘要
根据动车组维修的实际业务需求和应用环境,构建以事件为中心的动车组维修信息系统数据处理架构,实现对无线射频识别数据中高层业务信息的有效提取。从事件驱动的角度将数据处理结果分成原始事件、简单事件、复杂事件和应用服务4种形态。首先通过过滤校验将无线射频识别装置及软件触发产生的原始事件转换为简单事件;然后通过构建的复杂事件检测Petri网进行模式识别,从简单事件中发现复杂事件;最后以工序作业进度辅助监控、流水线瓶颈检测、员工作业轨迹跟踪、部件物流轨迹跟踪4个高层业务为例,由复杂事件构建应用服务。应用实践表明,该技术实现了高层业务信息的有效提取,提高了信息系统数据处理的效率与能力,解决了数据有余而信息不足的问题。
According to the practical business requirement and application environment of EMU （Electric Multiple Units） maintenance,an event-centered data processing framework for EMU maintenance information system was proposed to implement high-level information extraction from radio frequency identification data.The processing results were divided into four forms from event-driven perspective,including raw event,simple event,complex event and application service.Firstly,the raw events collected from radio frequency identification devices as well as software triggers were transformed into simple events through filtering mechanism.Secondly,the complex event detection Petri-net was established to perform pattern recognition for discovering complex events from simple events.Finally,with four high-level business examples of job progress monitoring,production line bottleneck detection,staff activity tracking and parts logistics tracking,application services were built based on complex events.The practice shows that this technology has realized the effective extraction of high-level information,improved the efficiency and capacity of data processing and solved the problem of sufficient data whereas insufficient information.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2014年第4期108-116,共9页
China Railway Science
基金
铁道部科技研究开发计划项目(2011J002)
