DocumentCode :
3399899
Title :
The Harmonic Assessment Method Based on the Generalized Ridge Regression
Author :
Yao Shu-jun ; Yan, Wang ; Wang Yan ; Liu Long-hui ; Ma Lu-yao
Author_Institution :
North China Electr. Power Univ., Beijing, China
fYear :
2012
fDate :
27-29 March 2012
Firstpage :
1
Lastpage :
4
Abstract :
This article will use the generalized ridge regression in system harmonic impedance analysis and the harmonic evaluation. Based on the equivalent circuit of the system, users can obtain system harmonic impedance, the background harmonic and common connection point, the analytical expression of the harmonic. Using the harmonic voltage and current data which will be got in the common connection point, people can use system harmonic impedance and background harmonic as the regression coefficients of generalized ridge regression. Using direct method can obtain estimation parameters of the generalized ridge regression, and calculates the harmonic emission levels. The method can overcome the influence caused by the abnormal data to Regression coefficient, it is better than the partial least squares method and its ridge estimators have better robustness. The simulation results prove the effectiveness of this method and accuracy.
Keywords :
equivalent circuits; power system harmonics; regression analysis; background harmonic; common connection point; equivalent circuit; estimation parameters; generalized ridge regression; harmonic assessment method; harmonic emission levels; harmonic evaluation; harmonic voltage; partial least squares method; regression coefficients; ridge estimators; system harmonic impedance analysis; Equivalent circuits; Estimation; Harmonic analysis; Impedance; Least squares methods; Mean square error methods; Power system harmonics;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Power and Energy Engineering Conference (APPEEC), 2012 Asia-Pacific
Conference_Location :
Shanghai
ISSN :
2157-4839
Print_ISBN :
978-1-4577-0545-8
Type :
conf
DOI :
10.1109/APPEEC.2012.6307682
Filename :
6307682
Link To Document :
بازگشت