摘要
随着下一代互联网的快速发展,云网运营面临前所未有的挑战,包括服务的高度复杂性、广泛的数据处理要求和对实时响应的严格要求。针对这些问题,研究了云网运维智能化转型的创新实践,包括基于DevOps的故障识别、智能推荐系统的故障定位、大模型辅助的故障处置、人机交互的故障调度以及智能分类的工单质检。基于创新实践应用,重点讨论了Fuzzy Matching与BERT结合的智能推荐算法、大模型在故障处置中的应用,以及多维度工单智能质检技术。
With the rapid development of the next generation Internet,the maintenance of cloud network is facing unprecedented challenges,including the high complexity of services,extensive data processing requirements and strict requirements for realtime response.In response to these problems,innovative practices for the intelligent transformation of cloud network operations and maintenance are studied,including fault recognition based on DevOps,fault localization of intelligent recommendation systems,fault handling assisted by large models,fault scheduling through human-machine interaction,and intelligent classification of work order quality inspection.Based on innovative practical applications,it focuses on the intelligent recommendation algorithm combined with Fuzzy Matching and BERT,the application of LLM(Large Language Model)in fault handling,and the multi-dimensional work order intelligent quality inspection technology.
作者
严劲
刘珮
Yan Jin;Liu Pei(China Telecom Corporation Limited,Beijing 100010,China)
出处
《邮电设计技术》
2024年第4期26-31,共6页
Designing Techniques of Posts and Telecommunications
关键词
云运维
智能化转型
大模型
故障处置
Cloud network operation and maintenance
Intelligent transformation
LLM
Fault handling
