基于深度学习的桥梁状态监测设备故障自动容错系统  

Automatic Fault Tolerance System of Bridge Condition Monitoring Equipment Based on Deep Learning

作　　者：梁洪永 李镇 袁梦雄 顾义 郝作锐 LIANG Hongyong;LI Zhen;YUAN Mengxiong;GU Yi;HAO Zuorui(Zhejiang Huadong Engineering Consulting Co.Ltd,Hangzhou 310014,China;College of Environment and Civil Engineering,Chengdu University of Technology,Chengdu 610059,China)

机构地区：[1]浙江华东工程咨询有限公司,杭州310014 [2]成都理工大学环境与土木工程学院,成都610059

出　　处：《自动化与仪器仪表》2021年第11期96-99,共4页Automation & Instrumentation

基　　金：国家自然科学基金面上项目(基于细观结构的砂卵石隧道围岩力学行为及拱效应机理研究,51978088,2020/01-2023/12)。

摘　　要：桥梁状态监测设备故障容错效率低会影响桥梁安全使用,造成隐患。因此,设计基于深度学习的桥梁状态监测设备故障自动容错系统。在软件方面,设计故障容错系统流程结构框架,提取桥梁状态监测设备故障特征,采用深度学习方法,对桥梁状态监测设备故障融合进行自适应学习和跟踪训练。在硬件方面,利用物联网组网技术获得双端口的PCI协议,在PCI总线控制协议的基础上,进行桥梁状态监测设备故障容错诊断的多通道联合控制,对寄存器初始化处理,实现桥梁状态监测设备故障自动容错系统的硬件构建。测试结果表明,设计的桥梁状态监测设备故障自动容错系统输出可靠性较好,故障容错收敛能力较好,有效提高桥梁状态监测能力。Low fault tolerance efficiency of bridge condition monitoring equipment will affect the safe use of bridge and cause hidden trouble.Therefore,an automatic fault tolerance system for bridge condition monitoring equipment based on deep learning is designed.In the aspect of software,the process structure framework of fault tolerance system is designed,the fault characteristics of bridge condition monitoring equipment are extracted,and the deep learning method is adopted to carry out adaptive learning and tracking training for the fault fusion of bridge condition monitoring equipment.In terms of hardware,the dual port PCI protocol is obtained by using the networking technology of the Internet of things.Based on the PCI bus control protocol,the multi-channel joint control of fault tolerance diagnosis of bridge condition monitoring equipment is carried out,and the register is initialized to realize the hardware construction of automatic fault tolerance system of bridge condition monitoring equipment.The test results show that the designed automatic fault tolerance system of bridge condition monitoring equipment has good output reliability,good fault tolerance convergence ability,and effectively improves the bridge condition monitoring ability.

关 键 词：深度学习 桥梁状态监测 故障自动容错系统 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TN366[自动化与计算机技术—控制科学与工程]