Dépôt numérique

Quasi-Monte Carlo technique in global sensitivity analysis of wind resource assessment with a study on UAE.


Téléchargements par mois depuis la dernière année

Plus de statistiques...

Tsvetkova, Olga; Ouarda, Taha B. M. J. (2019). Quasi-Monte Carlo technique in global sensitivity analysis of wind resource assessment with a study on UAE. Journal of Renewable and Sustainable Energy , vol. 11 , nº 5. 053303. DOI: 10.1063/1.5120035.

[img] PDF
Document sous embargo jusqu'à 8 octobre 2020.

Télécharger (4MB)


The present paper bridges mathematical modeling and wind resource assessment (WRA). Sensitivity analysis (SA) links portions of output variance to the variance in each input variable. Global SA (GSA) explores inputs globally. One-at-a-time SA is dominating in WRA, while GSA is often overlooked. Compared to traditional methods, GSA offers potential improvement by the means of the quasi-Monte Carlo (QMC) technique with its elaborate sampling designs enabling faster convergence. The main novelty of this work is twofold: the use of QMC in WRA and Sobol method as a variance-based GSA method in WRA. This paper is among a few using GSA in WRA. Two case studies were conducted. One shows that QMC with sampling based on Sobol sampling outperforms Latin hypercube sampling and pseudorandom sampling. It also shows that in terms of accuracy of results, the brute-force calculation of Sobol sensitivity indices (SI) should be used whenever the model allows it; otherwise, SI can be estimated. Another case study demonstrates a valid GSA study for WRA at Masdar City, United Arab Emirates. The results suggest that the influence of the variance in Weibull parameters on annual energy production (AEP) might be overestimated, while found to be responsible for 2% of AEP uncertainty, and the influence of the variance in air density might be overlooked, while found to account for 94%. WRA studies would benefit greatly from using the QMC and Sobol method. The Sobol method is a universal GSA method, providing valid results for nonlinear models typical for WRA, and QMC provides global scalability, model independence, and flexibility in uncertainty quantification.

Type de document: Article
Mots-clés libres: sampling; sensitivity analysis; wind power
Centre: Centre Eau Terre Environnement
Date de dépôt: 17 avr. 2020 18:33
Dernière modification: 17 avr. 2020 18:33
URI: http://espace.inrs.ca/id/eprint/9701

Actions (Identification requise)

Modifier la notice Modifier la notice