Induced voltage prediction in transmission line parameter testing based on hyperparameter optimized Set-Transformer

The analysis and prediction of induced voltages in transmission lines are of great significance for ensuring the safety of line operation and maintenance. To address the varying coupling effects caused by electromagnetic environment non-uniformity in non-fully parallel lines, a novel prediction method based on an improved Set Transformer is proposed. First, a feature set characterizing line-induced voltages is extracted through theoretical analysis based on historical test data. Hyperparameters of the Set Transformer are then optimized by the Optuna optimizer, and an Optuna-ST based prediction model is established for both electrostatically and electromagnetically induced voltages. The model is trained and tested on the collected samples, and its effectiveness is further validated through field case studies. The results show that the Optuna-ST model can achieve a Normalized Mean Absolute Error (NMAE) of about 6%, a Tolerance accuracy above 85%, and a coefficient of determination (R²) exceeding 0.98. The model can accurately predict induced voltages between non-fully parallel transmission lines. Compared with representative baselines such as Convolutional Neural Network (CNN) and Light Gradient Boosting Machine (LightGBM), this method has higher prediction accuracy, reliability, and generalization capability, providing a crucial decision-making basis for ensuring safety during line parameter testing.

Authors

Institutions

Publication Details

Journal
Archives of Electrical Engineering
Published
2026-11-18
DOI
https://doi.org/10.24425/aee.2026.11175
Primary Topic
High-Voltage Power Transmission Systems
Type
article
Field-Weighted Citation Impact
0.00

Funders

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

Induced voltage prediction in transmission line parameter testing based on hyperparameter optimized Set-Transformer

Peng Chang, Liangliang Song, Chenqing Wang
1 links
Archives of Electrical Engineering
High-Voltage Power Transmission Systems
article

Induced voltage prediction in transmission line parameter testing based on hyperparameter optimized Set-Transformer

Peng Chang, Liangliang Song, Chenqing Wang
article en

Abstract

The analysis and prediction of induced voltages in transmission lines are of great significance for ensuring the safety of line operation and maintenance. To address the varying coupling effects caused by electromagnetic environment non-uniformity in non-fully parallel lines, a novel prediction method based on an improved Set Transformer is proposed. First, a feature set characterizing line-induced voltages is extracted through theoretical analysis based on historical test data. Hyperparameters of the Set Transformer are then optimized by the Optuna optimizer, and an Optuna-ST based prediction model is established for both electrostatically and electromagnetically induced voltages. The model is trained and tested on the collected samples, and its effectiveness is further validated through field case studies. The results show that the Optuna-ST model can achieve a Normalized Mean Absolute Error (NMAE) of about 6%, a Tolerance accuracy above 85%, and a coefficient of determination (R²) exceeding 0.98. The model can accurately predict induced voltages between non-fully parallel transmission lines. Compared with representative baselines such as Convolutional Neural Network (CNN) and Light Gradient Boosting Machine (LightGBM), this method has higher prediction accuracy, reliability, and generalization capability, providing a crucial decision-making basis for ensuring safety during line parameter testing.

Archives of Electrical Engineering
Electric Power Research Institute (US), Shanghai Electric (China) (CN), State Grid Jiangsu Electric Power (China) (CN)
Science and Technology Project of State Grid, State Grid Jiangsu Electric Power
Industry, innovation and infrastructure, Affordable and clean energy
High-Voltage Power Transmission Systems
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