Research Area


AI-Based Power
Infrastructure Diagnostics

Go →

Smart Grid
Big Data Analytics

Go →

Energy Storage
System Management

Go →

AI-Based Power Infrastructure Diagnostics

Reflectometry-Based Cable Diagnostics

  • As dependence on power cables is increasing correspondingly with skyrocketing demand for electric power every year, securing the reliability of power cables is important.
  • In this research, the reliability of the power system is improved through the development of cable diagnosis and monitoring techniques based on time-frequency domain reflectometry.
  • This research includes extending cable diagnostics to intelligent information technology (diagnosis automation, fault classification) by means of artificial intelligence.
Research Highlights
  • Diagnosis on high temperature superconducting (HTS) cables
  • Condition assessment of instrumentation and control (I&C) cables in nuclear power plants
  • Fault localization technique on HVDC/submarine cables
▲ Applications to actual field (KEPCO substations)
▲ Machine learning-based time-frequency domain reflectometry *
* C.-K. Lee, and Y.-J. Shin, “Detection and Assessment of I&C Cable Faults Using Time-Frequency R-CNN Based Reflectometry,” IEEE trans. Ind. Electron., vol. 68, no. 2, pp. 1581-1590, Feb. 2020.

Condition-Based Cable Health Index

  • An equipment having poor reliability can cause deadly failures and have corresponding adverse effects on the entire power system. Therefore, the condition of the equipment should be checked and maintained to determine marginal replacement time.
  • The goal of this research is to prevent cable failures and accidents in advance by indexing the condition of cables based on operational data such as cable load and failure histories.
  • The influences of thermal and mechanical stress on the lifespan of cables are investigated to propose a health index that reflects the overall complex stress, which will aid in securing the reliability of the cables.
Research Highlights
  • Development of intelligent health index of cable that reflects operational data in real time.
  • Investigation on the relationship between thermal and mechanical stress that affects the lifespan of cables and the proposed health index.
▲ Neural network-based cable health index algorithm *
▲ Cable temperature estimation using thermal circuit-based cable modeling **
* G. H. Ji, S. S. Bang, Y. H. Jung, T. I. Jang, and Y.-J. shin, “Ensemble Learning-Based Health Index for Underground Transmission Line,” Submitted to IEEE Trans. Ind. Informat.

** KEPCO KDN Research Project, “Development of Dynamic Rating System and Health Indexing for Distribution Cable.”

Extended Applications of Reflectometry

  • Diagnostic techniques in various fields have limitations in their performance because the diagnostic techniques analyze signals in the time domain or the frequency domain only.
  • Therefore, the research aims to improve the performance of the conventional diagnostic techniques through the convergence between the conventional diagnostic techniques and time-frequency domain reflectometry, which shows excellent performance in power cable diagnostics.
  • By analyzing the signal propagation characteristics of the diagnosis target, the type of diagnosis signal is determined, and research on the signal selection and application method is conducted.
  • Ongoing research includes ultrasonic-based reflectometry for pipeline diagnosis and laser-based reflectometry for optical cable diagnosis.
Research Highlights
  • Ultrasonic guided wave-based time-frequency domain reflectometry for diagnosis of pipelines
  • Intensity modulated laser-based time-frequency domain reflectometry for diagnosis of optical cables
▲ Guided wave-based time-frequency domain reflectometry *
* S. S. Bang, Y. H. Lee, and Y.-J. Shin, "Defect Detection in Pipelines via Guided Wave Based Time-Frequency Domain Reflectometry," IEEE Trans. Instrum. Meas., vol 70, pp. 9505811, Jan. 2021.

COPYRIGHT ⓒ 2021 연세대학교 에너지 정보처리 연구실. ALL RIGHTS RESERVED. design by dsso