A TLS-ESPRIT Method Based Fourth-Order Mixed Mean Cumulant Applying for Identiﬁcation of Subsynchronous Oscillation Modes in High Voltage DC System

doi:10.11985/2017.12.005

Abstract

Abstract:

Besides white Gaussian noise, there is also colored Gaussian noise in the collected signals of the subsynchronous oscillations because the signals have been processed by band-pass filters. To identify modes rapidly and accurately, a method is proposed which combine fourth-order mixed mean cumulant with total least squares-estimation of signal parametersvia rotational invariance techniques (TLS-ESPRIT) algorithm in the paper. Considered the effect of colored Gaussian noise as well as white Gaussian noise , the proposed method substitudes the sampled signals by its fourth-order mixed mean cumulant (FOMMC) sequence along the diagonal line. The simulation result shows the robustness of the proposed method which can identify the dominant modes from the signals with additive colored and white Gaussian noise, and it helps to suppress subsynchronous oscillations further.

Key words: Subsynchronous oscillations, fourth-order mixed mean cumulant, TLS-ESPRIT, white Gaussian noise, colored Gaussian noise

CLC Number:

TM712

Hu Nan. A TLS-ESPRIT Method Based Fourth-Order Mixed Mean Cumulant Applying for Identiﬁcation of Subsynchronous Oscillation Modes in High Voltage DC System[J]. Journal of Electrical Engineering, 2017, 12(12): 33-39.

Figures/Tables 13

Fig.1

Fig.2

Tab.1

Tab.2

Tab.3

Tab.4

Fig.3

Tab.5

Fig.4

Tab.6

Fig.5

Fig.6

Tab.7

References 21

[1]	Prabha Kunder. 电力系统稳定与控制[M]. 北京: 中国电力出版社, 2002.
[2]	李兴源 . 高压直流输电系统[M]. 北京: 科学出版社, 2010.
[3]	程时杰, 曹一家, 江全元 . 电力系统次同步振荡的理论与方法[M]. 北京: 科学出版社, 2009.
[4]	IEEE Subsynchronous Resonance Task Force. First bench-mark model for computer simulation of subsynchronous resonance[J]. IEEE Trans. on Power Apparatus System, 1977,96(5):1565-1572. doi: 10.1109/T-PAS.1977.32485
[5]	倪以信, 陈寿孙, 张宝霖 . 动态电力系统的理论和分析[M]. 北京: 清华大学出版社, 2002.
[6]	Girgis A A, Ham F M . A quantitative study of pitfalls in the FFT[J]. IEEE Transactions on Aerospace and Electronic Systems, 1980,16(10):434-439.
[7]	Lin Xiangning, Zhang Hao, Liu Pei , et al. Wavelet based scheme for detection of torsional osicallion[J]. IEEE Transactions on Power Systems, 2002,17(4):1096-1101. doi: 10.1109/TPWRS.2002.804962
[8]	李天云, 谢家安, 张方彦 , 等. HHT 在电力系统低频振荡模态参数提取中的应用[J]. 中国电机工程学报, 2007,27(28):79-83.
	Li Tianyun, Xie Jiaan, Zhang Fangyan , et al. Application of HHT for extracting model parameters of low frequency oscillations in power systems[J]. Proceedings of the CSEE, 2007,27(28):79-83.
[9]	Trudnowski D J . Making PRONY analysis more accurate using multiple signals[J]. IEEE Trans. on Power Systems, 1999,14(1):226-231. doi: 10.1109/59.744537
[10]	Hauer J F . Application of Prony analysis to the determination of modal content and equivalent models for measured power system response[J]. IEEE Trans. on Power Systems, 1991,6(3):1062-1068. doi: 10.1109/59.119247
[11]	Kumaresan M R, Tufts D W , et al. Estimating the parameters of exponentially damped sinusoids and pole-zero modeling in noise[J]. IEEE Trans. on Acoustics, Speech and Signal Processing, 1982,30(6):833-840. doi: 10.1109/TASSP.1982.1163974
[12]	王宇静, 于继来 . 电力系统振荡模态的矩阵束辨识法[J]. 中国电机工程学报, 2007,27(19):12-17.
	Wang Yujing, Yu Jilai . Matrix pencil method of oscillation modes identification in power systems[J]. Proceedings of the CSEE, 2007,27(19):12-17.
[13]	Rouquette S, Najim M . Estimation of frequencies and damping factors by two-dimensional ESPRIT type methods[J]. IEEE Trans. on Signal Processing, 2001,49(1):237-245. doi: 10.1109/78.890367
[14]	Ottersten B, Viberg M, Kailath T . Performance analysis of the total least squares ESPRIT algorithm[J]. IEEE Trans. on Signal Processing, 1991,39(5):1122-1135. doi: 10.1109/78.80967
[15]	张敏, 沈健, 侯国明 , 等. 相量测量单元实现次同步振荡在线辨识和告警的探讨[J]. 电力系统自动化, 2016,16(40):143-152.
	Zhang Min, Shen Jian, Hou Mingguo , et al. Discussion on on-line identification and waring of subsynchronous oscillation for phasor measuring unit[J]. Automation of Electronic Power System, 2016,16(40):143-152.
[16]	Giannikis G B, Zhou G . Harmonics in multiplicative and additive noise: parameter estimation using cyclic statistics[J]. IEEE Trans. on Signal Processing, 1995,43(9):2217-2221. doi: 10.1109/78.414790
[17]	Anderson M M, Giannakis G B, Swmai A . Harmonic retrieval using higher order statisties: adeterministic formulation[J]. IEEE Trans. on Signal Processing, 1995,43(8):1880-1889. doi: 10.1109/78.403347
[18]	Ruiz D P, Carrion M C, Gallego A , et al. Parmaeter estimation of exponentially damped sinusoids using a higher order correlation-based approaeh[J]. IEEE Trans. on Signal Processing, 1995,43(11):2665-2677. doi: 10.1109/78.482116
[19]	Dandwate A V, Giannakis G B . Asymptotie theory of mixed time means and kth-order cyclic-moment and cumulant statisties[J]. IEEE Trans. on IT, 1995,41(1):216-232. doi: 10.1109/18.370106
[20]	胡楠, 李兴源, 李宽 , 等. 基于四阶混合平均累计量的矩阵束算法在低频振荡在线辨识中的应用[J]. 中国电机工程学报, 2013,3(5):53-60.
	Hu Nan, Li Xingyuan, Li Kuan , et al. A modified matrix pencil method based on fourth—order mixed mean cumulantfor low frequency oscillation mode on–line identificatio[J]. Proceedings of the CSEE, 2013,3(5):53-60.
[21]	杨秀, 王西田, 陈陈 . 基于H_∞鲁棒控制理论的高压直流输电系统附加次同步振荡阻尼控制设计[J]. 电网技术, 2006,30(9):57-61.
	Yang Xiu, Wang Xitian, Chen Chen . Design of supplemental subsynchronous oscillation damping control based on H_∞ robust control theory of HVDC system[J]. Power System Technology, 2006,30(9):57-61.

频率/Hz	衰减系数	误差(%)
14.999 3	-0.033 8	12.7
31.999 6	-0.203 2	1.6
45.999 5	-0.003 3	34

频率/Hz	衰减系数	误差(%)
14.999 9	-0.029 5	1.67
31.999 9	-0.198 6	0.70
45.999 8	-0.005 6	12

频率/Hz	衰减系数	误差(%)
15.000 0	-0.029 9	0.33
31.999 9	-0.200 3	0.15
46.000 0	-0.004 8	4

辨识算法	频率/Hz	误差(%)	衰减系数	误差(%)
Prony	15.002 5	0.017	-0.024 5	27.5
	32.002 1	0.006 6	-0.280 2	23.1
	46.001 1	0.002 4	-0.009 9	98
MP	15.000 6	0.004	-0.032 7	9
	32.001 7	0.005 3	-0.219 9	9.95
	46.000 2	0.000 4	-0.007 1	42
TLS-ESPRIT	15.000 5	0.003 3	-0.027 6	8
	31.998 5	0.004 7	-0.180 3	9.85
	45.999 7	0.000 6	-0.002 8	44
FOMMC-TLS-ESPRIT	15.000 0	0	-0.029 9	0.33
	32.000 0	0	-0.199 9	0.05
	45.999 9	0	-0.004 7	6

辨识算法	频率/Hz	衰减系数	阻尼比(%)
MP	13.383 9	0.079 0	-0.094
	24.433 0	0.056 4	-0.037
	29.505 2	-0.008 0	0.005
TLS-ESPRIT	13.385 5	0.080 5	-0.096
	24.434 5	0.057 7	-0.037
	29.508 6	-0.009 2	0.005
FOMMC-TLS-ESPRIT	13.393 1	0.079 5	-0.095
	24.433 3	0.058 1	-0.038
	29.503 2	-0.008 4	0.005