Abstract

For sustained low frequency oscillations in steady-state power systems, an algorithm is proposed for precise online identification of oscillation frequency, oscillation amplitude, and fundamental amplitude. The algorithm consists of a robust low frequency estimator and a notch filter in parallel. The asymptotical convergence property is analyzed by slow integral manifold, averaging method, and Lyapunov stability theorem sequentially. The steady-state antinoise property is investigated by perturbed system theorem. The robust advantages of the proposed algorithm are embodied in the following aspects: the fundamental amplitude identification is little influenced by oscillation frequency and oscillation amplitude, both oscillation frequency identification and oscillation amplitude identification have small steady-state errors under high order harmonics or bounded noises, the ramp variations of both fundamental amplitude and oscillation amplitude are also significantly tracked, and three design parameters have different effects on identification performance of oscillation frequency, oscillation amplitude, and fundamental amplitude, respectively. Simulation results verify the validity.

Highlights

  • Low-frequency oscillations (LFOs) in power system are usually in range of 0.1-2.5Hz and may result in unstable and unsecure operation, so its real-time identification is essential [1,2,3]

  • Consider some measured voltage signals or current signals with sustained low frequency oscillation in steady-state power systems expressed in time t as y (t) = y1 (t) + yo (t) + n (t) (1)

  • If γ=0 the estimated frequency θ is frozen and the state equation (9) becomes a Four-dimension linear time invariant (LTI) dynamic system excited by two sinusoidal signal

Read more

Summary

Introduction

Low-frequency oscillations (LFOs) in power system are usually in range of 0.1-2.5Hz and may result in unstable and unsecure operation, so its real-time identification is essential [1,2,3]. An inherent disadvantage of the estimators proposed in [7, 9, 11,12,13] is that the low frequency convergence property conflicts to the high frequency convergence speed This was investigated and modified in [14] by a robust technique to precisely identify very low frequency. In this paper, for the sustained LFOs, we employ the robust low frequency estimator proposed in [14] in parallel to a notch filter for certain frequency sinusoid to online identify the oscillation frequency, the oscillation amplitude, and the fundamental amplitude.

Framework of the Proposed Algorithm
Convergence Property Analysis
Antinoise Property Analysis
Simulation Results
Summary
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call