摘要
根据桥梁实地健康监测数据,分析其海量数据的特点并阐述了采用数据驱动对其进行处理的必要性和优越性。研究基于数据驱动的桥梁健康监测信息预处理方法,借鉴小波分析法和自适应控制法建立模型。经对实桥挠度数据的处理仿真结果表明,该方法可以有效地在监测信息存入数据库前将原始信息进行分类统计,寻找出异常数据和缺失数据后,将异常数据进行分离,将缺失数据进行补偿,进而提高后期数据处理效率。将处理得到的数据经高低频分离后,其瞬变信息刚好反映了结构在动荷载作用下的挠度变化情况。
According to the characteristics of bridge health monitoring data,illustrate the importance of using data-driven to process the mass data. Research the method of preprocessing bridge health monitoring data based on data-driven,and build a model by reference of wavelet analysis method and adaptive control method. The analysis of experimental results shows the method can process the original in-formation by statistical analysis to classify. It will separate the singular data and compensate the missing data,after finding out the abnor-mal and missing data. And further increase the later data processing efficiency. After separating the data by high＆low frequency,it shows that the transient information just reflects the structure in the dynamic loads deflection changes.
出处
《计算机技术与发展》
2013年第10期258-260,F0003,共4页
Computer Technology and Development
基金
交通运输部
西部交通建设科技项目(20113188141480)
关键词
数据驱动
桥梁健康监测
数据处理
海量数据
data-driven
bridge health monitoring
data processing
massive data
