摘要
针对我国移动通信互联终端用户越来越多,造成通信网络故障率较高,且检测难度大的问题,设计了一个基于数字孪生技术在移动通信中的故障检测系统。利用数字孪生技术实现对移动通信的各应用层创建数字世界;通过持续监控真实物理系统并使用大数据分析和机器学习来预测现实世界中发生的故障问题;利用生成对抗网络算法对转化到数字孪生技术内的移动通信数据进行计算检测,从而改善持续运营的问题。另外通过长短期记忆网络算法对整个故障识别模块进行改进,利用长短期记忆网络算法对历史数据智能存储的特点,达到对整个通信网络故障特征的提取,提高故障检测的效率和准确度。试验结果表明,系统技术核算的数据、误差率在可接受范围内,为其他技术研究奠定基础。
A fault detection system based on digital twin technology in mobile communication was designed to address the increasing number of mobile communication interconnection terminal users in China,which led to a high failure rate and difficulty in detection of communication networks.The digital twin technology was utilized to create a digital world for various application layers of mobile communication;by continuously monitoring real physical systems and using big data analysis and machine learning,faults that occur in the real world were predicted;the generative adversarial network algorithm was used to compute and detect mobile communication data transformed into digital twin technology,thereby improving the problem of continuous operation.In addition,the entire fault recognition module was improved through the long short-term memory(LSTM)network algorithm,which can intelligently store historical data and extract fault features of the entire communication network to improve the efficiency and accuracy of fault detection.The experimental results indicate that the data error rate of the system technical accounting is within an acceptable range,laying the foundation for other technical research.
作者
南作用
钟志刚
陈任翔
王亚
Nan Zuoyong;Zhong Zhigang;Chen Renxiang;Wang Ya(China Information Consulting&Designing Institute Co.,Ltd.,China Unicom,Zhengzhou Henan 450000,China)
出处
《电气自动化》
2024年第3期108-112,共5页
Electrical Automation
关键词
数字孪生技术
生成对抗网络算法
长短期记忆算法
故障识别
故障检测
digital twin(DT)technology
generative adversarial network(GAN)algorithm
long short-term memory(LSTM)algorithm
fault identification
failure detection