NDE reliability gains from combining eddy-current and ultrasonic testing

Hörn, D.; Mayo, W.R

doi:10.1016/s0963-8695(99)00058-4

Cited by 55 publications

(28 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The traditional application area for data fusion has long been target detection and tracking [38], [39], [50]- [54]. However, the aforementioned approaches, with their many variations and combinations, are becoming increasingly popular for emerging areas, such as remote sensing [36], [45], [55], [56], biometrics [57], [58], nondestructive evaluation (NDE) [59], [60], electroencephalographic data analysis [61], speech processing [62], [63], information retrieval [49], [64], and many others. Excellent review articles on various data fusion methods can be found in [50] and [65]- [67].…”

Section: B Ensemble Approaches and Data Fusionmentioning

confidence: 99%

An Ensemble-Based Incremental Learning Approach to Data Fusion

Parikh

Polikar

2007

IEEE Trans. Syst., Man, Cybern. B

View full text Add to dashboard Cite

Abstract-This paper introduces Learn++, an ensemble of classifiers based algorithm originally developed for incremental learning, and now adapted for information/data fusion applications. Recognizing the conceptual similarity between incremental learning and data fusion, Learn++ follows an alternative approach to data fusion, i.e., sequentially generating an ensemble of classifiers that specifically seek the most discriminating information from each data set. It was observed that Learn++ based data fusion consistently outperforms a similarly configured ensemble classifier trained on any of the individual data sources across several applications. Furthermore, even if the classifiers trained on individual data sources are fine tuned for the given problem, Learn++ can still achieve a statistically significant improvement by combining them, if the additional data sets carry complementary information. The algorithm can also identify-albeit indirectly-those data sets that do not carry such additional information. Finally, it was shown that the algorithm can consecutively learn both the supplementary novel information coming from additional data of the same source, and the complementary information coming from new data sources without requiring access to any of the previously seen data.Index Terms-Data fusion, incremental learning, Learn++, multiple classifier/ensemble systems.

show abstract

Section: B Ensemble Approaches and Data Fusionmentioning

confidence: 99%

An Ensemble-Based Incremental Learning Approach to Data Fusion

Parikh

Polikar

2007

IEEE Trans. Syst., Man, Cybern. B

View full text Add to dashboard Cite

show abstract

“…The probability of detecting a given deterioration state (event D) is a conditional probability, that is, Prob(D|X(t) = k, α), in so far as the capability of the inspection technique depends, in general, on the degree of deterioration present, say, X(t) = k-in general, a higher degree of deterioration is more likely to be detected than only minor deterioration-and the amount of effort spent in assessing the structural condition. For example, utilizing different, partial complementary inspection methods 51) may lead to an overall improved inspection capability, described herein by the parameter of inspection quality α, but may also result in higher inspection costs I(α). It should go without saying, that the inspection efforts can vary from maintenance to maintenance, that is, for each maintenance at time t j we can have a different parameter α j of inspection quality.…”

Section: (2) Quality Of Inspectionmentioning

confidence: 99%

“…Thereby, rehabilitation work is always preceded by inspection at cost I(α j ), which is a function of the parameter α j of inspection quality. That is, a more detailed inspection, utilizing maybe different condition assessment procedures 45), 51) , is more costly than a routine visual inspection. Since inspection or condition assessment makes only sense as long as the structure has not failed, the cost for inspection I(α j ) has to multiplied by the probability of no fail-…”

Section: (3) Expected Lossesmentioning

confidence: 99%

Cost-Benefit Based Optimization of Maintenance Interventions for Deteriorating Structures

Higuchi

Macke

2007

Structural Eng./Earthquake Eng.

View full text Add to dashboard Cite

In most more economically developed countries an ever growing percentage of existing structures is threatened by obsolescence in the short-to medium-term-either because of structural deficits due to deterioration, or due to functional aging. To ensure sustained serviceability and safety of these structures, maintenance interventions are utilized, which allow partial or complete structural rehabilitation. However, such maintenance interventions have to be economically reasonable, that is, maintenance expenditures spent have to be outweighed by expected future benefits. For this purpose, we propose herein a novel optimization formulation for maintenance planning based on cost-benefit criteria. The usefulness of the proposed approach lies in the fact, that it not only allows to determine optimal sequences of maintenance times, rehabilitation levels and inspection qualities, but also allows to specify economically optimal lifetimes and acceptable failure rates of structures. The modeling of structural deterioration and maintenance, as well as the setting of all relevant cost factors is discussed in detail. Numerical examples investigate the effect of imperfect execution of maintenance actions and functional aging.

show abstract

“…However, this will increase the cost and time for inspection, and the time taken to analyse results from all techniques separately. This problem can be minimised by combining more than one technique into a single probe, data acquisition system and analysis algorithm, significantly reducing inspection time [2][3][4][5].…”

Section: Introductionmentioning

confidence: 99%

“…Improvements to NDT can also be made when using two or more different techniques which are sensitive to similar types of defect [2][3][4][5]. When data analysis is done concurrently, an increased reliability of information about the defects can be given by competitive data fusion.…”

Section: Introductionmentioning

confidence: 99%

Data fusion for defect characterisation using a dual probe system

Edwards

Sophian

Dixon

et al. 2008

Sensors and Actuators A: Physical

View full text Add to dashboard Cite

We present recent work on a dual probe system containing electromagnetic acoustic transducers (EMATs) generating and detecting surface ultrasonic waves, and a pulsed eddy current (PEC) probe. This system is able to detect and size surface and near-surface defects in electrically conducting samples by looking at changes in the detected signal for each probe. By combining the information from each probe using a weighted logic function for data fusion, it is possible to both classify and size defects, with increased reliability. By combining the data in this way one obtains information about the defects which is not available when using either probe in isolation. Typical results on steel and aluminium samples are presented, along with information about the data fusion function. The dual probe and data fusion routine has been demonstrated to work at manualscanning speeds, with higher speeds possible following some simple improvements to the system.

show abstract

NDE reliability gains from combining eddy-current and ultrasonic testing

Cited by 55 publications

References 3 publications

An Ensemble-Based Incremental Learning Approach to Data Fusion

An Ensemble-Based Incremental Learning Approach to Data Fusion

Cost-Benefit Based Optimization of Maintenance Interventions for Deteriorating Structures

Data fusion for defect characterisation using a dual probe system

Contact Info

Product

Resources

About