Direction of arrival estimation method for multiple UHF partial discharge sources based on virtual array extension

• Published in: IEEE transactions on dielectrics and electrical insulation Vol. 25; no. 4; pp. 1526 - 1534
• Main Authors: Zhou, Nan, Luo, Lingen, Sheng, Gehao, Jiang, Xiuchen
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.08.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Arrays; compressive sensing; Computer simulation; Direction of arrival; Direction-of-arrival estimation; DOA estimation; Estimation; Focusing; Matching pursuit algorithms; Mathematical analysis; Matrix algebra; Matrix methods; multiple partial discharge; Multiple signal classification; Partial discharges; Pitch (inclination); Sensor arrays; Signal reconstruction; spatial spectrum; UHF; Very high frequencies; virtual array extension
• ISSN: 1070-9878; 1558-4135
• DOI: 10.1109/TDEI.2018.007093

The localization method of multiple partial discharge (PD) of power equipment in airinsulated substation based on traditional ultra-high frequency (UHF) sensor array suffer from the problem of low accuracy, due to the lack of enough sensors. Thus, a novel virtual array extension method based on Compressive Sensing (CS) is proposed in this paper. The traditional 2×2 UHF array is virtually extended to a N×N (N>2) array, with the steering vector matrix as the basis transformation matrix, the random Gaussian matrix as the measurement matrix, and the orthogonal matching pursuit (OMP) algorithm as the signal reconstruction algorithm. The PD signal received by the extended array is used to construct the PD spatial spectrum, on the basis of wide-band signal focusing. The directional angle and pitch angle of PD sources are obtained by performing peak search of the PD spatial spectrum. The simulation results show that the extended 4×4 array can detect up to 3 PD sources simultaneously. Moreover, the directional of arrival (DOA) estimation accuracy is greatly improved. Compared to the traditional 2×2 array, the directional angle estimation error is reduced from 10° to 4°. The experiment results further verified the effectiveness of the proposed method.