Modes preserving wavelet based multi-scale PCA algorithm for compression of smart grid data
Authors:
- 1 August 2013
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 817-821
Abstract
In this paper, wavelet based multiscale PCA algorithm is proposed for effective compression of smart grid data under normal as well as fault conditions. The signal decomposition is done using wavelet transform, followed by the de-correlation using PCA, to achieve maximum compression, while simultaneously preserving the dominant modes of the signal and bad data rejection. The optimum decomposition scale of wavelet transform is selected based on the energy of wavelet coefficients in each scale. The proposed algorithm employing multiscale PCA computes the principal components of the wavelet coefficients at each scale, followed by combining the results at relevant scales. The wavelet coefficients of a particular scale corresponding to the dominant eigen values are retained for signal compression. The dominant modes in the signal are sorted based on their energy content. This overcomes the drawback of false detection of modes and lower accuracy of estimation which arises when systems of higher order are used. Prony analysis is performed to check ability of the compression strategy to preserve the modal information. The phasor data are simulated under fault conditions in the IEEE 30-bus system. The results from Prony analysis indicate that multiscale PCA effectively compresses the disturbance signals while preserving the model information of the signal.Keywords
This publication has 10 references indexed in Scilit:
- Smart Grid — The New and Improved Power Grid: A SurveyIEEE Communications Surveys & Tutorials, 2011
- Principal component analysis based compression scheme for power system steady state operational data2011
- A Wavelet-Based Data Compression Technique for Smart GridIEEE Transactions on Smart Grid, 2010
- Compressing Electrical Power Grids2010
- Multiscale Windowed Denoising and Segmentation of Hyperspectral Images2008
- Model identification and error covariance matrix estimation from noisy data using PCAControl Engineering Practice, 2008
- Multivariate calibration of non-replicated measurements for heteroscedastic errorsChemometrics and Intelligent Laboratory Systems, 2007
- Edge detection by scale multiplication in wavelet domainPattern Recognition Letters, 2002
- Multiscale PCA with application to multivariate statistical process monitoringAIChE Journal, 1998
- A new wavelet-based edge detector via constrained optimizationImage and Vision Computing, 1997