作　　者： ; ; ;

机构地区： 中兴通讯股份有限公司

出　　处： 《中兴通讯技术》 2017年第4期2-8,共7页

摘　　要： 通过对中兴通讯大视频运维系统整体架构和关键模块的介绍,以及机器学习技术在大视频运维系统中端到端异常检测、根因分析与故障预测等场景的具体应用的分析,并结合硬盘故障预测的实例,认为随着人工智能在运维领域的应用发展,从基于规则的自动化运维转向基于机器学习的智能运维必然成为趋势。中兴通讯适时采用了机器学习方法来提取历史巡检数据中蕴含的故障特征,并构建集成预测模型来提升大视频运维的精度和效率,目前取得了较好的效果。 In this paper,the overall architecture and key modules of ZTE big video operation and maintenance(O&M)system are introduced.Then the application scenes of machine learning technology in this system,including the end-to-end anomaly detection,root cause analysis and fault prediction are analyzed,as well as the instances of hard disk breakdown prediction.With the development of artificial intelligence technologies in the field of O&M,the trend which is from rule-based automatic to machine learning-based intelligent will be formed.To improve the precision and efficiency of the big video O&M,ZTE has adopted a machine learning method to extract the fault features contained in the historical inspection data,and has built the integrated prediction model,which has achieved good performances.

关 键 词： 大视频 大数据 机器学习 人工智能

领　　域： []