Civil and Environmental Engineering Department Seminar: Harnessing Data Analytics and Computational Models in Structural Monitoring
Seminar | February 17 | 12-1 p.m. | 542 Davis Hall
In recent years, advances in informatics and data science have assisted engineers to tackle structural dynamics problems. For example, health monitoring of structure and infrastructure systems has become a successful paradigm, as a valuable source of information for evaluating structural integrity and reliability throughout the lifecycle of structures as well as ensuring optimal maintenance planning and operation. Important development in sensor, computer and data analytics technologies made it possible to process big amount of data, to mine characteristic features, and to link those to the current structural conditions. In this presentation, I will talk about harnessing data analytics and computational models to tackle structural monitoring issues, through signal processing, identification, inference, computational modeling and uncertainty quantification. This talk will mainly discuss a typical topic on combined data analytics and computational models for building monitoring to show the basic concept. Deconvolution interferometry is employed for processing the vibration data, extracting wave propagation information and thus identifying structural characteristics. The extracted waves are then used for parameter uncertainty quantification of a computational model within the framework of hierarchical Bayesian inference. The presented methodologies can be used to process and mine big monitoring data for a real time operating system, and show a great potential in assessing structural integrity leading to a smart structure management system.