TY - JOUR
T1 - Two-Level Ambient Oscillation Modal Estimation From Synchrophasor Measurements
JF - IEEE Transactions on Power Systems
Y1 - 2014/12//
SP - 1
EP - 10
A1 - Ning, Jiawei
A1 - S. Arash Nezam Sarmadi
A1 - Venkatasubramanian, Vaithianathan
KW - AA13-004
KW - AARD
KW - CERTS
KW - oscillations
KW - phasor measurement units (PMUs)
AB - This paper proposes a decentralized two-level structure for real-time modal estimation of large power systems using ambient synchrophasor data. It introduces two distributed algorithms that fit the structure well, namely, 1) decentralized frequency domain decomposition and 2) decentralized recursive stochastic subspace identification. As opposed to present-day oscillation monitoring methodologies, the bulk of the algorithmic computations is done locally at the substation level in the two-level framework. Substation modal estimates are sent to the control center where they are grouped, analyzed, and combined to extract system modal properties of local and inter-area modes. The framework and the proposed algorithms provide a scalable methodology for handling oscillation monitoring from a large number of substations efficiently. The two-level structure and the two decentralized algorithms are tested using simulated data from standard test systems and from archived real power system synchrophasor data.
JO - IEEE Trans. Power Syst.
DO - 10.1109/TPWRS.2014.2373172
ER -
TY - JOUR
T1 - Oscillation modal analysis from ambient synchrophasor data using distributed frequency domain optimization
JF - IEEE Transactions on Power Systems
Y1 - 2013/05//
SP - 1960
EP - 1968
A1 - Ning, Jiawei
A1 - Pan, Xueping
A1 - Venkatasubramanian, Vaithianathan
KW - AA13-004
KW - AARD
KW - CERTS
KW - oscillations
KW - phasor measurement units (PMUs)
AB - This paper provides a distributed frequency domain algorithm for real-time modal estimation of large power systems using ambient synchrophasor data. By dividing the computation between a supervisory central computer and local optimizations at the substation level, the algorithm efficiently estimates multiple dominant mode frequencies, damping ratios and mode shapes from wide-area power system measurements. The algorithm, called distributed frequency domain optimization, is tested on known test systems and archived real power system data from eastern and western power systems.
VL - 28
IS - 2
JO - IEEE Trans. Power Syst.
DO - 10.1109/TPWRS.2013.2248028
ER -