A Bayesian approach to optimal sensor placement for structural health monitoring with application to active sensing

Flynn, Eric B.; Todd, Michael D.

doi:10.1016/j.ymssp.2009.09.003

Cited by 215 publications

(147 citation statements)

References 13 publications

Supporting

Mentioning

147

Contrasting

Order By: Relevance

“…(2), a statistical model could be alternatively employed that incorporates the probability of damage occurring at any given pixel location, as well as the associated costs of false-detects and misses. This is the general approach taken by Flynn and Todd [8,9]; however, their work effectively focuses on the pixel value when damage is present and does not consider imaging artifacts.…”

Section: Pixel-specific Performance Ratiomentioning

confidence: 99%

“…The concept of establishing pixel-specific thresholds for damage detection, which was recently proposed by Flynn and Todd [8,9], represents a departure from previous guided wave imaging performance metrics that are focused on image quality. These metrics typically characterize the performance of an array by comparing the pixel value at one or more potential damage locations to the pixel values at all non-damage locations for a single image.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Analysis of Distributed Sparse Array Configurations for Guided Wave Imaging Applications

Hall¹,

Michaels²

2011

AIP Conference Proceedings

View full text Add to dashboard Cite

ABSTRACT. Permanently attached, distributed sparse arrays of piezoelectric transducers have been proposed as a cost-effective solution for rapid interrogation of large, plate-like structures for structural health monitoring. Many proposed methods using these arrays rely on guided wave imaging techniques to interpret and graphically display information for damage detection and localization. Guided wave imaging algorithms, however, contain imaging artifacts that are due to reflections from boundaries and other structural features, even in a noise-free environment. The magnitude relationship between pixel values corresponding to actual damage and imaging artifacts is dependent on a number of variables, including imaging algorithm, number and location of transducers in the array, location and scattering behavior of damage, and structural geometry. This study proposes a metric to characterize the imaging performance of distributed sparse array systems and utilizes numerical simulations to analyze the impact of damage location, imaging algorithm, and physical array configuration on performance in the context of a single flaw in a rectangular aluminum plate.

show abstract

Section: Pixel-specific Performance Ratiomentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Analysis of Distributed Sparse Array Configurations for Guided Wave Imaging Applications

Hall¹,

Michaels²

2011

AIP Conference Proceedings

View full text Add to dashboard Cite

show abstract

“…Ferentinos and Tsiligiridis 15 proposed a multi-objective optimization strategy to implement in the wireless sensor networks via GA. Kang et al 16 investigated three-sensor configuration performance indexes and proposed a virus co-evolutionary partheno-genetic algorithm. Flynn and Todd 17 proposed a novel Bayesian scheme to carry out optimal sensor placement which could be applicable on active sensing as well. Chen and Nagarajaiah 18 presented a new method according to the GA to identify the detection filter-based decentralized controller in which numerical and laboratory results demonstrated that the strategy can detect structural damage.…”

Section: Introductionmentioning

confidence: 99%

Optimal placement of sensors and piezoelectric friction dampers in the pipeline networks based on nonlinear dynamic analysis

Shokouhi

Yuan

Zhu

2017

Journal of Low Frequency Noise, Vibration and Active Control

View full text Add to dashboard Cite

Experiences of past earthquakes demonstrate that pipeline systems have no proper performance when exposed to severe earthquakes. In this study, sensor and damper placement approaches are presented for doing reliable health monitoring and seismic retrofitting of the piping networks. Since most of the available sensor placement methods are based on modal analysis results, the authors propose a new scheme that relies on the nonlinearity which utilizes nonlinear time history analysis results, and genetic algorithm is selected to act as the methodology of optimization as well. The results demonstrate that the proposed optimal sensor configuration strategy is more accurate and efficient than the extended modal assurance criterion method. To assess the number of sensors, a sensitivity analysis is undertaken in which the number of sensors computed optimally by the proposed algorithm contains the least convergence error. In addition, the number of iterations and the time consumed in the proposed approach are considerably less than the extended modal assurance criterion method. Moreover, the efficiency of the proposed sensor placement scheme was compared with a new algorithm proposed by Sun and Büyükö ztürk, named discrete artificial bee colony, where the simulation result demonstrates high accuracy of the proposed sensor configuration approach. The initial time history analysis results show the vulnerable points of the system, which destroyed due to the applied seismic waves. Hence, to enhance the seismic performance of the system, piezoelectric friction dampers are optimally placed, where it can be clearly seen that the optimal arrangement of piezoelectric friction dampers in the piping system can significantly decrease the seismic response.

show abstract

“…Flynn and Todd [5] proposed a Bayesian approach for determining the optimal sensor placement by minimizing the Bayes risk of false positive and false negative detection error. Thiene et al [6] developed an algorithm to obtain the optimal sensor network with the maximum area coverage.…”

Section: Introductionmentioning

confidence: 99%

Damage Detectability Model of Pitch-Catch Configuration in Composite Plates

Yue

Khodaei

2017

KEM

View full text Add to dashboard Cite

Abstract. Detectability of damage using Lamb waves depends on many factors such as size and severity of damage, attenuation of the wave and distance to the transducers. This paper presents a detectability model for pitch-catch sensors configuration for structural health monitoring (SHM) applications. The proposed model considers the physical properties of lamb wave propagation and is independent of damage detection algorithm, which provides a generic solution for probability of detection. The applicability of the model in different environmental and operational conditions is also discussed.

show abstract

A Bayesian approach to optimal sensor placement for structural health monitoring with application to active sensing

Cited by 215 publications

References 13 publications

Analysis of Distributed Sparse Array Configurations for Guided Wave Imaging Applications

Analysis of Distributed Sparse Array Configurations for Guided Wave Imaging Applications

Optimal placement of sensors and piezoelectric friction dampers in the pipeline networks based on nonlinear dynamic analysis

Damage Detectability Model of Pitch-Catch Configuration in Composite Plates

Contact Info

Product

Resources

About