Integrated infrastructure health monitoring : detection, estimation, and learning
Abstract/Contents
- Abstract
- Buildings, bridges, power grids, and other infrastructure are essential parts of the urban system. It is critical to diagnose their health conditions continuously. The availability of low-cost sensing, wireless communications, and data processing technologies has made it possible to deploy large-scale sensor networks to monitor urban infrastructure. The data from these networks offer opportunities to integrate spatial and temporal information for infrastructure health monitoring. Effective detection and localization of damage require models and inference algorithms that combine sensor data, knowledge of the underlying system dynamics, and the spatial and temporal aspects of measurements. In the first part of this dissertation, we design statistical and machine learning based methods to recover spatial and temporal representations of the data. For example, graphical models are utilized to model dependencies between infrastructure damage and sensor measurements. The methods can also recover the underlying infrastructure network and its connectivity. We illustrate the approach by demonstrating a provably correct method to recover the distribution power grid topology from power measurements at the network endpoints. The algorithm is tested on a large-scale data set of power measurements from a local utility. In the second part of this dissertation, we focus on designing algorithms for detection and localization of damage. Practically useful detection algorithms utilize a stream of data to decide about damage state continuously. Change point detection algorithms utilize the fact that the statistical distribution of the data changes before and after damage. When these distributions are known, optimal algorithms that minimize the delay in detecting damage for a given false alarm level can be designed. When the post-damage distribution is not known, some form of inference needs to be performed simultaneously with detection. In addition, when multiple types or locations of damage are present, then multiple change points need to be considered. We propose utilizing graphical models to create algorithms that can learn the post-damage distributions and cooperatively detect multiple change points leading to localizing damage in space and time. A cooperative Bayesian damage detection and localization method is shown to isolate the damage location in an optimal way and can be even computed in a distributed manner. The methods are tested utilizing various datasets collected from experiments with buildings subject to different ground motions in shake table tests. Then, we transfer the learning in structural damage detection to identify power line outage in distribution grid. We demonstrate the same data-driven method can solve problems in two different physical systems. We conclude the dissertation by presenting a novel wireless sensor system, SnowFort, that is designed for infrastructure health monitoring, with the integration of decision making system.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2018; ©2018 |
| Publication date | 2018; 2018 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Liao, Yizheng |
|---|---|
| Degree supervisor | Rajagopal, Ram |
| Thesis advisor | Rajagopal, Ram |
| Thesis advisor | Baker, Jack W |
| Thesis advisor | Kiremidjian, Anne S. (Anne Setian) |
| Degree committee member | Baker, Jack W |
| Degree committee member | Kiremidjian, Anne S. (Anne Setian) |
| Associated with | Stanford University, Civil & Environmental Engineering Department. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Yizheng Liao. |
|---|---|
| Note | Submitted to the Civil & Environmental Engineering Department. |
| Thesis | Thesis Ph.D. Stanford University 2018. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2018 by Yizheng Liao
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...