A magnetic characteristics health assessment framework for synchronous generators based on three-phase dynamic Potier reactance and machine learning

With the increasing penetrations of renewable energy, turbo-generators are required to operate under deeper peak-shaving conditions, which accelerates the degradation of magnetic characteristics. Magnetic characteristics of synchronous generators can be assessed from two perspectives: the overall magnetic field level and the inter-phase magnetic balance. However, existing methods either rely on offline no-load tests or lack physical interpretability when based on operating data. This paper proposes a magnetic characteristics health assessment framework for synchronous generators based on three-phase dynamic Potier reactance and machine learning. The proposed model characterises magnetic saturation and distortion under varying operating conditions using measured electrical quantities. On this basis, two types of magnetic characteristic analysis are achieved. First, the dynamic Potier reactance model is incorporated into a conventional generator model to calculate the offset of the no-load characteristic. A Fourier neural network with residual learning is then employed to reconstruct the aged characteristic curve, enabling online assessment of magnetic ageing. Second, inter-phase variations of the estimated Potier reactance are analysed to evaluate magnetic balance, enabling online detection and classification of rotor eccentricity faults. The proposed framework is evaluated using coupled FEM–circuit simulations and data from a 644 MW turbo-generator. The simulations validate the ageing and eccentricity reference models, while the measured data validate aged-characteristic reconstruction and non-eccentric-condition identification.

Authors

Institutions

Publication Details

Journal
Electric Power Systems Research
Published
2026-11-16
DOI
https://doi.org/10.1016/j.epsr.2026.114487
Primary Topic
Electric Motor Design and Analysis
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A magnetic characteristics health assessment framework for synchronous generators based on three-phase dynamic Potier reactance and machine learning

Blair Brown, Tong Jiang, Yongjin Xu, Mulin Liu
Electric Power Systems Research
Electric Motor Design and Analysis
article

A magnetic characteristics health assessment framework for synchronous generators based on three-phase dynamic Potier reactance and machine learning

Blair Brown, Tong Jiang, Yongjin Xu, Mulin Liu
article en

Abstract

With the increasing penetrations of renewable energy, turbo-generators are required to operate under deeper peak-shaving conditions, which accelerates the degradation of magnetic characteristics. Magnetic characteristics of synchronous generators can be assessed from two perspectives: the overall magnetic field level and the inter-phase magnetic balance. However, existing methods either rely on offline no-load tests or lack physical interpretability when based on operating data. This paper proposes a magnetic characteristics health assessment framework for synchronous generators based on three-phase dynamic Potier reactance and machine learning. The proposed model characterises magnetic saturation and distortion under varying operating conditions using measured electrical quantities. On this basis, two types of magnetic characteristic analysis are achieved. First, the dynamic Potier reactance model is incorporated into a conventional generator model to calculate the offset of the no-load characteristic. A Fourier neural network with residual learning is then employed to reconstruct the aged characteristic curve, enabling online assessment of magnetic ageing. Second, inter-phase variations of the estimated Potier reactance are analysed to evaluate magnetic balance, enabling online detection and classification of rotor eccentricity faults. The proposed framework is evaluated using coupled FEM–circuit simulations and data from a 644 MW turbo-generator. The simulations validate the ageing and eccentricity reference models, while the measured data validate aged-characteristic reconstruction and non-eccentric-condition identification.

Electric Power Systems ResearchVol. 265
North China Electric Power University (CN), University of Strathclyde (GB), State Nuclear Power Technology Company (China) (CN)
China Power International Development Limited
Affordable and clean energy, Industry, innovation and infrastructure
Electric Motor Design and Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon