基于人工智能的运营商故障分析能力提升研究

Research on Improving Fault Analysis Capability of Operators Based on Artificial Intelligence

传统的故障分析手段将运维经验固化为故障分析规则或脚本,这种方式针对特定故障模式较为有效,但是无法应对新网络业务、组网变化,一旦规则、脚本需要调整,需投入较大成本进行适配改造,且时效性较差。而人工智能技术在大数据统计、分析、推理、自适应学习上有着先天优势,能快速基于新网络、新业务的变化重训练AI模型参数,给出最佳推荐值。基于此,重点研究了如何利用人工智能提升运营商故障分析能力。

Traditional fault analysis methods are fixed into fault analysis rules or scripts based on O&M experience.This method is effective for specific fault modes,but cannot cope with new network service and networking changes.Once these rules and scripts need to be adjusted,a large cost is required for adaptation and reconstruction,and the timeliness is poor.Artificial intelligence technology has inherent advantages in big data statistics,analysis,inference,and adaptive learning.It can quickly retrain AI model parameters based on changes in new networks and businesses,and provides the best recommendation value.Based on this,it focus on the research of using artificial intelligence to improve the fault analysis capability of operators.