Compressive sensing with an adaptive wavelet basis for structural system response and reliability analysis under missing data

Comerford, Liam; Jensen, Héctor A.; Mayorga, F.; Beer, Michael; Kougioumtzoglou, Ioannis A.

doi:10.1016/j.compstruc.2016.11.012

Cited by 38 publications

(16 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compressive sampling (or sensing, CS) is a novel sampling paradigm in digital signal processing to reconstruct a signal (e.g., an image with 1000 × 1000 = 1 million pixels) from a small number of measurements on that signal (Candès et al 2006;Donoho 2006;Candès and Wakin 2008;Wang and Zhao 2016;Comerford et al 2016Comerford et al , 2017. In the context of signal processing, Table 3 is a 9 × 9 matrix and can be considered as an image with 9 × 9 = 81 pixels and missing values at 9 pixels.…”

Section: Compressive Samplingmentioning

confidence: 99%

Managing Risk in Geotechnical Engineering – From Data to Digitalization

Phoon¹,

Ching²,

Yu³

2019

Proceedings of the 7th International Symposium on Geotechnical Safety and Risk (ISGSR 2019)

View full text Add to dashboard Cite

If you scan a page from a soil report, this is called digitization. If you deploy digital technologies, both software such as building information modeling and machine learning and hardware such as autonomous drones and additive manufacturing, to support new and more collaborative forms of project delivery, this is called digitalization. Data lies at the heart of this transformation that is targeted at re-valuing infrastructure from a "brick and mortar" asset to a service for the interests of the end-users. There is a need to view the value of data completely differently from how they are routinely used in current practice. In particular, there is a need to treat data as assets in themselves, over and above their conventional roles as inputs to a physical model or as monitoring data to trigger interventions. This paper explores the availability and nature of geotechnical data and presents two recent advances made in this direction for a specific but important task of estimating soil/rock properties (compressive sampling and Bayesian machine learning). Data-driven decision making does not imply taking the engineer out of the entire life cycle management chain. It is intended to support rather than to replace human judgment.

show abstract

Section: Compressive Samplingmentioning

confidence: 99%

Managing Risk in Geotechnical Engineering – From Data to Digitalization

Phoon¹,

Ching²,

Yu³

2019

Proceedings of the 7th International Symposium on Geotechnical Safety and Risk (ISGSR 2019)

View full text Add to dashboard Cite

show abstract

“…For robust transmission and recovery of structural images, Yang and Nagarajaiah 19 proposed a CS-based method that preserved essential information for structural health diagnosis while reduced the data size notably. Comerford et al 20 used CS with an adaptive wavelet basis to handle the missing data problem and then conducted structural reliability analysis. Other methods have also been investigated for SHM data completion.…”

Section: Introductionmentioning

confidence: 99%

Group sparsity-aware convolutional neural network for continuous missing data recovery of structural health monitoring

Tang

Bao

2020

Structural Health Monitoring

View full text Add to dashboard Cite

In structural health monitoring, data quality is crucial to the performance of data-driven methods for structural damage identification, condition assessment, and safety warning. However, structural health monitoring systems often suffer from data imperfection, resulting in some entries being unusable in a data matrix. Discrete missing points are relatively easy to recover based on known adjacent points, whereas segments of continuous missing data are more common and also more challenging to recover in a practical scenario. Formulating the data recovery task as an optimization problem for matrix completion, we present a convolutional neural network to achieve simultaneous recovery for multi-channel data with the awareness of group sparsity. The data recovery process based on compressive sensing is formulated as a regression problem and achieved in the neural network. The basis matrix is utilized as the input and the incomplete data matrix as the output to provide partial information for approximation. Basis coefficient optimization is performed via convolutional operation. Group sparsity regularization is applied while updating the kernel of the convolutional layer. The recovery can be readily obtained after optimization (training) without further validation and testing. The proposed method does not need intact data prepared in advance for training; also, it can handle sporadic data loss and make the most of interrupted information. Recovery ability evaluations on synthetic data, field-test data, and monitoring data of seismic response indicate that the proposed method achieves a good recovery result with high loss ratio and continuous data loss. The code is available at https://github.com/dawnnao/Group-sparsity-aware-CNN .

show abstract

“…extreme features or other causes such as sensor maintenance, bandwidth limitation or data acquisition restrictions could lead to poor quality of the data records. For this reason, the real data records must be represented in an appropriate manner and the uncertainties mitigated as much as possible [2].…”

Section: Introductionmentioning

confidence: 99%

“…In addition, the underlying physics should be understood well. Because both cases are often not satisfied, other approaches must be found to develop a load model which represents the data in the best possible way [2].…”

Section: Introductionmentioning

confidence: 99%