铁路人像检索平台流量智能调度研究  

Intelligent traffic dispatching of railway portrait retrieval platform

作　　者：李贝贝[1] 朱建生 阎志远 戴琳琳 解敏森 LI Beibei;ZHU Jiansheng;YAN Zhiyuan;DAI Linlin;XIE Minsen(Postgraduate Department,China Academy of Railway Sciences,Beijing 100081,China;Institute of Computing Technologies,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)

机构地区：[1]中国铁道科学研究院研究生部,北京100081 [2]中国铁道科学研究院集团有限公司电子计算技术研究所,北京100081

出　　处：《铁路计算机应用》2021年第8期6-10,共5页Railway Computer Application

基　　金：中国铁道科学研究院集团有限公司科研项目(2019YJ122)。

摘　　要：铁路人像检索平台建立之初,对业务地址的管理是通过Apollo配置中心手动修改的方式完成。操作流程复杂,运维效率低下。流量智能导航可以有效解决业务地址的均衡调度问题。文章描述人像检索平台的功能组成与技术架构,指出既有平台的运维效率低下与流量调度不均衡等问题;通过配置改造,形成业务路由地址池,并以各车站预估的日客流量作为参考,提出车站的等级划分,形成车站权重值;设计依托权重的流量智能调度策略。业务流量智能调度的实现,简化了运维工作,使得人像检索平台的业务流量调度更加均衡。At early stage of the establishment of the railway face retrieval platform,the management of business address was completed by the manual modification of Apollo configuration center.Its operation process was complicated and maintenance efficiency was low.The intelligent traffic navigation could effectively solve the problem of balanced dispatching with the business address.This paper described the functional composition and technical architecture of the portrait retrieval platform,pointed out the problems of the existing platform such as low operating efficiency and unbalanced traffic dispatching.Through configuration transformation,the service routing address pool was formed,and according to the estimated daily passenger flow of each station,the concept of station grade was put forward and the station weight was formed,intelligent traffic dispatching strategy was designed based on weight.The implementation of intelligent traffic dispatching simplified the work of operation and maintenance,made the traffic dispatching of portrait retrieval platform more balance.

关 键 词：铁路人像检索 Apollo配置 流量智能调度 导航 电子客票 

分 类 号：U293.22[交通运输工程—交通运输规划与管理]