Abstract
A statistical-based approach to detect outliers in fluid-based velocity measurements is proposed. Outliers are effectively detected from experimental unimodal distributions with the application of an existing multivariate outlier detection algorithm for asymmetric distributions (Hubert and Van der Veeken, J Chemom 22:235–246, 2008). This approach is an extension of previous methods that only apply to symmetric distributions. For fluid velocity measurements, rejection of statistical outliers, meaning erroneous as well as low probability data, via multivariate outlier rejection is compared to a traditional method based on univariate statistics. For particle image velocimetry data, both tests are conducted after application of the current de facto standard spatial filter, the universal outlier detection test (Westerweel and Scarano, Exp Fluids 39:1096–1100, 2005). By doing so, the utility of statistical outlier detection in addition to spatial filters is demonstrated, and further, the differences between multivariate and univariate outlier detection are discussed. Since the proposed technique for outlier detection is an independent process, statistical outlier detection is complementary to spatial outlier detection and can be used as an additional validation tool.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
MATLAB® 7.9.0.529 (R2009b) The MathWorks, Inc. Natick, MA.
DaVis 7.4.0.93. LaVision, Inc. Goettingen, Germany.
In between passes of multiple pass algorithms, DaVis 7.4 automatically fills holes with interpolated values and applies a 3 × 3 smoothing filter.
References
Barnett V, Lewis T (1984) Outliers in statistical data. Wiley, New York
Becker C, Gather U (2001) The largest nonidentifiable outlier: a comparison of multivariate simultaneous outlier identification rules. Comput Stat Data Anal 36:119–127. doi:10.1016/S0167-9473(00)00032-3
Bendat JS, Piersol AG (2000) Random data: analysis and measurement procedures, 3rd edn. Wiley-Interscience, New York
Brys G, Hubert M, Struyf A (2004) A robust measure of skewness. J Comput Graph Stat 3:996–1007
Coleman HW, Steele WG (1999) Experimentation and uncertainty analysis for engineers. Wiley, New York
Gnanadesikan R (1997) Methods for statistical data analysis of multivariate observations. Wiley, New York
Hadi A (1992) Identifying multiple outliers in multivariate data. J R Stat Soc B Stat Methodol 54:761–771
Hawkins DM (1980) Identification of outliers. Chapman and Hall, New York
Hubert M, Vandervieren E (2008) An adjusted boxplot for skewed distributions. Comput Stat Data Anal 52:5186–5201. doi:10.1016/j.csda.2007.11.008
Hubert M, Van der Veeken S (2008) Outlier detection for skewed data. J Chemom 22:235–246. doi:10.1002/cem.1123
Keane R, Adrian R (1990) Optimization of particle image velocimeters. Part I: double-pulsed systems. Meas Sci Technol 1:1202–1215. doi:10.1088/0957-0233/1/11/013
Mahalanobis P (1936) On the generalized distance in statistics. Proc Natl Acad Sci India A 2:49–55
Maronna R, Yohai V (1995) The behavior of the Stahel-Donoho robust multivariate estimator. J Am Stat Assoc 90:330–341
Mathieu J, Scott J (2000) An introduction to turbulent flow. Cambridge University Press, New York
Murray NE, Ukeiley LS (2007) An application of Gappy POD—for subsonic cavity flow PIV data. Exp Fluids 42:79–91. doi:10.1007/s00348-006-0221-y
Nauri S, Nogueira J, Lecuona A, Legrand M, Rodriguez PA (2008) Limits and accuracy of the stereo-lfc piv technique and its application to flows of industrial interest. Exp Fluids 45(4):609–621. doi:10.1007/s00348-008-0545-x
Raffel M, Willert C, Wereley S, Kompenhans J (2007) Particle image velocimetry: a practical guide. Springer, Berlin
Rocke D, Woodruff D (1996) Identification of outliers in multivariate data. J Am Stat Assoc 91:1047–1061
Rousseeuw P, Van Driessen K (1999) A fast algorithm for the minimum covariance determinant estimator. Technometrics 41:212–223
Shinneeb A, Bugg J, Balachandar R (2004) Variable threshold outlier identification in PIV data. Meas Sci Technol 15:1722–1732. doi:10.1088/0957-0233/15/9/008
Song X, Yamamoto F, Iguchi M, Murai Y (1999) A new tracking algorithm of PIV and removal of spurious vectors using Delaunay tessellation. Exp Fluids 26:371–380. doi:10.1007/s003480050300
Thompson WR (1935) On a criterion for the rejection of observations and the distribution of the ratio of deviation to sample standard deviation. Ann Math Stat 6:214–219
Verboven S, Hubert M (2005) LIBRA: a MATLAB library for robust analysis. Chemom Intell Lab 75:127–136. doi:10.1016/j.chemolab.2004.06.003
Westerweel J (1994) Efficient detection of spurious vectors in particle image velocimetry data sets. Exp Fluids 16:236–247. doi:10.1007/BF00206543
Westerweel J, Scarano F (2005) Universal outlier detection for PIV data. Exp Fluids 39:1096–1100. doi:10.1007/s00348-005-0016-6
Wetzel D, Griffin J, Liu F, Cattafesta L (2009) An experimental study of circulation control on an elliptic airfoil. AIAA Paper 2009-4280. In: 39th AIAA fluid dynamics conference, San Antonio, TX
Acknowledgments
The authors would like to thank Drew Wetzel and Nik Zawodny for sharing experimental results. In addition, the authors would like to acknowledge and thank the Robust Statistics research group at the Katholieke Universiteit Leuven for providing the LIBRA MATLAB toolbox free to academic users.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Rights and permissions
About this article
Cite this article
Griffin, J., Schultz, T., Holman, R. et al. Application of multivariate outlier detection to fluid velocity measurements. Exp Fluids 49, 305–317 (2010). https://doi.org/10.1007/s00348-010-0875-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00348-010-0875-3