Reconstruction of Lightning-Induced Overvoltages on Overhead Distribution Lines Through Physics–Data Synergy

Partially known lightning excitation parameters and computationally expensive repeated electromagnetic-field and line-response simulations complicate reconstructing lightning-induced overvoltages on overhead distribution lines. We develop the Physics–Data Synergistic Lightning Overvoltage Network (PDS-LONet), linking lightning-parameter inversion, spatial-field prediction and field-to-line coupling within the prescribed simulation domain. Monitoring data from 10 kV lines only characterized measured waveforms. Models of lightning return strokes, electromagnetic field propagation, and field-to-line coupling provided supervised samples. In full mode, time–frequency features jointly estimate the peak lightning current, wavefront parameter, wave-tail parameter, and horizontal lightning strike distance. Grounding resistance is specified as a known condition for line voltage reconstruction. Four-parameter inversion yielded a mean coefficient of determination of 0.9968. Closed-loop waveform reconstruction yielded a normalized root mean square error of 3.92%; single cascaded inference took approximately 110 ms. Given only the peak current and strike distance, nearest-neighbor completion yielded a coefficient of determination of 0.981 for peak voltage prediction. The mean absolute percentage error was 5.80%, compared with 12.70% for the Rusck formula using the same peak current and strike distance inputs. Within the tested simulation workflow, physical-model samples, intermediate physical quantities, and boundary-condition modulation enable fast surrogate reconstruction.

Authors

Institutions

Publication Details

Published
2026-11-17
DOI
https://doi.org/10.22541/authorea.15012377/v1
Primary Topic
Lightning and Electromagnetic Phenomena
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Reconstruction of Lightning-Induced Overvoltages on Overhead Distribution Lines Through Physics–Data Synergy

Qing Yang, Jing Chang, Tao Wei, Chenjie Zhang et al.
Lightning and Electromagnetic Phenomena
preprint

Reconstruction of Lightning-Induced Overvoltages on Overhead Distribution Lines Through Physics–Data Synergy

Qing Yang, Jing Chang, Tao Wei, Chenjie Zhang, Tong Zhou, Simin Ye
preprint en

Abstract

Partially known lightning excitation parameters and computationally expensive repeated electromagnetic-field and line-response simulations complicate reconstructing lightning-induced overvoltages on overhead distribution lines. We develop the Physics–Data Synergistic Lightning Overvoltage Network (PDS-LONet), linking lightning-parameter inversion, spatial-field prediction and field-to-line coupling within the prescribed simulation domain. Monitoring data from 10 kV lines only characterized measured waveforms. Models of lightning return strokes, electromagnetic field propagation, and field-to-line coupling provided supervised samples. In full mode, time–frequency features jointly estimate the peak lightning current, wavefront parameter, wave-tail parameter, and horizontal lightning strike distance. Grounding resistance is specified as a known condition for line voltage reconstruction. Four-parameter inversion yielded a mean coefficient of determination of 0.9968. Closed-loop waveform reconstruction yielded a normalized root mean square error of 3.92%; single cascaded inference took approximately 110 ms. Given only the peak current and strike distance, nearest-neighbor completion yielded a coefficient of determination of 0.981 for peak voltage prediction. The mean absolute percentage error was 5.80%, compared with 12.70% for the Rusck formula using the same peak current and strike distance inputs. Within the tested simulation workflow, physical-model samples, intermediate physical quantities, and boundary-condition modulation enable fast surrogate reconstruction.

Chongqing University (CN)
Industry, innovation and infrastructure
Lightning and Electromagnetic Phenomena
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