A New Method of Eliminating Outliers in Data Processing of Tracked Vehicle Vibration Signal

Abstract:

Article Preview

The existence of outliers can produce false harmonic frequency components in data processing for vibration signal of tracked vehicle.The paper uses the Wright standards to judge details coefficients of outliers and calculates new details coefficients to replace the ones of the outliers by weighted coefficient method.The signal can be reconstructed with new details coefficients .The results show that the new method can retain the maximum information of singal after eliminating the outliers , meet the data distribution and reduce variance.

Info:

Periodical:

Edited by:

Mohamed Othman

Pages:

317-320

Citation:

W. W. Lv et al., "A New Method of Eliminating Outliers in Data Processing of Tracked Vehicle Vibration Signal", Applied Mechanics and Materials, Vols. 229-231, pp. 317-320, 2012

Online since:

November 2012

Export:

Price:

$38.00

[1] Zhao S.Z., Gao C.W., Wu X.L. et. al. The research of eliminating the outlier in data processing of remote measurement[J]. Structure & environment engineering. 2005, 32(1): 59-63.

[2] Angiulli F, Basta S, Pizzuti C. Distance-Based Detection and Prediction of Outliers[C]. IEEE Transactions on Knowledge and Data Engineering, 2006: 145-160.

DOI: https://doi.org/10.1109/tkde.2006.29

[3] Sheng Y.J., Li Q.H., Li K.L. A New Approach for Mining Local Outlier[R]. Proceedings of the Second International Conference on Machine Learning and Cybernetics, Wan, 2003: 157-162.

[4] Li J.X., Wang Y., Wang S.Z., Li Q.M. Simulation and study of the methods distinguishing and dealing with outliers in observation[A]. Microcomputer Information Valuesvalues. 2006, 22(5): 140-142.

[5] Zang Y.P., Zhang D.J., Wang W.Z. Per-level threshold de-noising m ethod using wavelet and its application in engine vibration analysis[J]. Journal of vibration and shock, 2009, 28(8): 57-60.