摘要
针对老旧桥梁在监测过程中出现的难以诊断以及诊断结果不准确等问题,充分利用BP神经网络在分类识别上的优越性,创新性地提出了一种基于BP神经网络的老旧桥梁诊断方法。首先,提出了基于BP神经网络的系统总体方案,并利用MIDAS/Civil软件对桥梁进行建模仿真,获取桥梁震动样本数据;其次,搭建BP神经网络模型,并将桥梁数据的固有频率和位移作为网络的输入集,健康状况作为目标标签进行输出;最终,在MATLAB仿真环境中进行仿真实验,对网络模型进行训练测试,输出桥梁健康状况。试验结果表明,该方法可以在获得少量数据的前提下,进行桥梁健康诊断时效果明显,可有效识别桥梁的健康状况,具有一定的工程价值。
In view of the problems of the old bridge in the process of monitoring,such as the difficulty of diagnosis and the inaccuracy of the diagnosis results,this paper makes full use of the advantages of BP neural network in classification and recognition,and puts forward a new diagnosis method of the old bridge based on BP neural network.First of all,the overall scheme of the system based on BP neural network is proposed,and the MIDAS/Civil software is used to model and simulate the bridge to obtain the vibration sample data of the bridge;secondly,the BP neural network model is built,and the natural frequency and displacement of the bridge data are taken as the input set of the network,and the health status is taken as the target label for output;finally,the simulation is carried out in the MATLAB simulation environment In the experiment,the network model is trained and tested to output the health status of the bridge.The test results show that this paper can get a small amount of data on the premise of bridge health diagnosis effect is obvious,can effectively identify the health status of the bridge,has a certain engineering value.
作者
李禹剑
李剑
辛伟瑶
Li Yujian;Li Jian;Xin Weiyao(North University of China,Taiyuan 030051,China)
出处
《国外电子测量技术》
2020年第2期19-22,共4页
Foreign Electronic Measurement Technology
基金
国家自然基金青年科学基金(61901419)
山西省面上青年资金(201801D221205)
山西省高校创新项目(201802083)
“十三五”装备预研兵器工业联合基金(6141B012904)
装备预研兵器装备联合基金(6141B021303)项目资助.
关键词
桥梁监测
BP神经网络
健康诊断
bridge monitoring
BP neural network
health diagnosis
