Ambient data-based online identification and location of frequency oscillations

Guowei Cai, Lei Xuan, Zhenglong Sun, Jiang Chao, Juri Belikov, Yoash Levron

Research output: Contribution to journalArticlepeer-review


Frequency oscillation results in all units within the system oscillating in the same phase and persistently, leading to significant damage to the system. It is important to remove the threat of frequency oscillation. However, existing methods rely on extracting and localizing oscillation parameters from transient oscillation data, failing to effectively identify potential oscillation risks within the system for proper corrective control. To fill this gap, this paper proposes a method for identifying and localizing frequency oscillations within random environmental data. Firstly, signal processing employs Morlet wavelet filtering to extract electrical quantity data in a single mode. Subsequently, a subspace dynamic mode decomposition algorithm is applied for parameter identification, enabling the detection and warning of weakly damped modes. In order to accurately locate the source of oscillations, a locating index is improved based on the energy flow theory, known as the dissipation area difference ratio, which enables the online evaluation of dissipation energy variation in generators within power systems operating under different modes. Through an arithmetic analysis conducted on both a 16-machine 5-zone system and an actual grid system, the effectiveness and accuracy of the proposed identification and localization method are demonstrated.

Original languageEnglish
Article number109843
JournalInternational Journal of Electrical Power and Energy Systems
StatePublished - Jun 2024


  • Energy flow
  • Frequency oscillation
  • Morlet wavelet filtering
  • Random data
  • SDMD

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering


Dive into the research topics of 'Ambient data-based online identification and location of frequency oscillations'. Together they form a unique fingerprint.

Cite this