Accurate and Reduced Order Identification of Propagation Function for Electromagnetic Transient Analysis of Cables

Miguel Cervantes, Ilhan Kocar, Jean Mahseredjian, Abner Ramirez

Research output: Journal article publicationJournal articleAcademic researchpeer-review

3 Citations (Scopus)

Abstract

This paper proposes a set of fitting features that enables identifying propagation function of cables accurately in the form of multi-delay rational transfer functions while maintaining reduced order of approximation. Frequency domain partitioning and adaptive weighting techniques are applied directly in phase domain for the evaluation of poles and residues simultaneously to ensure the precision of fitting of all entries including the low-magnitude off-diagonal elements. The objective is not only to obtain a precise fitting in phase domain but also to account for intrinsic modal decomposition so that integration errors in time domain are not magnified. The order of approximation is considerably reduced by post-processing the fitting using the balanced realization technique. When the proposed fitting approach is combined with a precise integration technique in time domain, it leads to accurate evaluation of transients as demonstrated in this paper, eliminating the spurious oscillations or numerical instabilities that may be encountered in the universal line model regardless of the integration or interpolation technique.

Original languageEnglish
Article number8784261
Pages (from-to)968-976
Number of pages9
JournalIEEE Transactions on Power Delivery
Volume35
Issue number2
DOIs
Publication statusPublished - 1 Apr 2020
Externally publishedYes

Keywords

  • Cable modeling
  • electromagnetic transients
  • frequency partitioning
  • model order reduction

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Accurate and Reduced Order Identification of Propagation Function for Electromagnetic Transient Analysis of Cables'. Together they form a unique fingerprint.

Cite this