Efficient global sensitivity analysis with correlated variables

DeCarlo, Erin C.; Mahadevan, Sankaran; Smarslok, Benjamin P.

doi:10.1007/s00158-018-2077-1

Cited by 12 publications

(3 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Spearman correlation coefficients assume linearity in the system, which is often not the case in practice. Sobol' indices allow for nonlinearity, but assume all parameters to be independent to identify the influence of each input parameter on the output [6,10,[58][59][60][61][62][63][64][65][66]. Correlation coefficients should ideally be established between input parameters [63,67].…”

Section: Correlation and Sensitivity Analysismentioning

confidence: 99%

“…It is logical to assume there will be significant correlations between quantitative, measured variables and the qualitative influence on how those variables are recorded. Emerging techniques have been proposed to account for dependant variables in SA, with varying success [66][67][68]. Incorporation with qualitative uncertainties also requires further research at this stage [6,58,61,66].…”

Section: Research Gapsmentioning

confidence: 99%

“…Emerging techniques have been proposed to account for dependant variables in SA, with varying success [66][67][68]. Incorporation with qualitative uncertainties also requires further research at this stage [6,58,61,66]. The risks in ignoring correlation in uncertainty propagation and SA were explored extensively by Groen [63,68].…”

Section: Research Gapsmentioning

confidence: 99%

See 2 more Smart Citations

Compound Uncertainty Quantification and Aggregation for Reliability Assessment in Industrial Maintenance

et al. 2023

View full text Add to dashboard Cite

The mounting increase in the technological complexity of modern engineering systems requires compound uncertainty quantification, from a quantitative and qualitative perspective. This paper presents a Compound Uncertainty Quantification and Aggregation (CUQA) framework to determine compound outputs along with a determination of the greatest uncertainty contribution via global sensitivity analysis. This was validated in two case studies: a bespoke heat exchanger test rig and a simulated turbofan engine. The results demonstrated the effective measurement of compound uncertainty and the individual impact on system reliability. Further work will derive methods to predict uncertainty in-service and the incorporation of the framework with more complex case studies.

show abstract