Lightning Potential Prediction for Transmission Lines in the Guanzhong Region Based on Deep Learning

DONG Lifan; LI Wei; NAN Zewen; CUI Xiaopeng; YANG Lanjun

doi:10.16516/j.ceec.2025-170

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Lightning Potential Prediction for Transmission Lines in the Guanzhong Region Based on Deep Learning

更新时间：2026-10-14

- Lightning Potential Prediction for Transmission Lines in the Guanzhong Region Based on Deep Learning
- Southern Energy Construction Vol. 13, Issue 3, Pages: 92-102(2026)
- 作者机构：
  
  1. 西安交通大学电工材料电气绝缘全国重点实验室,陕西,西安,710049
  2. 咸阳市气象局,陕西,咸阳,712000
  3. 国网陕西省电力有限公司电力科学研究院,陕西,西安,710100
- 作者简介：
- 基金信息：
- DOI：10.16516/j.ceec.2025-170
  CLC： TM7;S716
- Published：2026
- 稿件说明：
移动端阅览
董立凡, 李伟, 南泽文, 等. 基于深度学习的关中地区输电线路区域雷电潜势预测[J]. 南方能源建设, 2026, 13(3): 92-102.

.
董立凡, 李伟, 南泽文, 等. 基于深度学习的关中地区输电线路区域雷电潜势预测[J]. 南方能源建设, 2026, 13(3): 92-102. DOI： 10.16516/j.ceec.2025-170.

. DOI： 10.16516/j.ceec.2025-170.

摘要

目的

由于输电线路分布范围广、高度高，极易遭受雷击的危害。文章旨在制定有效的事前雷电防护措施，实现输电线路雷击灾害的全过程防护。

方法

以关中地区为例，筛选出33个与雷电活动高度相关的物理量作为特征，利用参数敏感性自适应（Parameter Sensitivity Adaptation，PSA）算法优化学习率、隐藏层节点数及正则化系数，并基于多变量长短时记忆网络（Long Short Term Memory

LSTM）构建输电线路区域雷电潜势预测模型，从而实现未来12 h输电线路区域雷电活动的预测。

结果

研究表明，修正

指数、Charba修正指数等33个大气物理量与雷电活动高度相关，可作为雷电潜势预测的特征。在模型超参数方面，通过PSA算法确定的最佳组合为：学习率2.7×10

−4

、隐藏层节点数29、正则化系数

4.1582

×10

−4

。此外，当批次样本量设定为16、序列长度设定为20时，模型可达到最小误差值。在此条件下，预测正确率为80.94%，命中率为71.3%。

结论

综合考虑输电线路的地理分布及其与气象探空站之间的距离，预测结果具有一定的有效性。

Abstract

Objective

Due to the wide distribution range and high height of transmission lines

they are highly susceptible to lightning strikes. This study aims to develop effective pre lightning protection measures and achieve full process protection against lightning strikes disaster on transmission lines.

Method

Taking the Guanzhong region as

an example

selected thirty-three physical quantities highly correlated with lightning activity as features

and used the parameter sensitivity adaptation (PSA) algorithm to optimize the learning rate

number of hidden layer nodes and regularization coefficient. Based on a multivariate long short term memory (LSTM) network

a lightning potential prediction model for transmission line areas was constructed to predict lightning activity in the next twelve hours.

Result

Research indicates that thirty-three atmospheric physical quantities

such as the modified

index and Charba modified index

are highly correlated with lightning activity and can serve as features for lightning potential prediction. In terms of model hyperparameters

the optimal combination determined through the PSA algorithm is: learning rate of 2.7×10

−4

number of hidden layer nodes of 29

and regularization coefficient of

4.1582

×10

−4

. Furthermore

when the batch size is set to 16 and the sequence length is set to 20

the model achieves the minimum error value. Under these conditions

the prediction accuracy is 80.94%

and the probability of detection (POD) is 71.3%.

Conclusion

Taking into account the geographical distribution of transmission lines and their distance from meteorological sounding stations

the prediction results have a certain degree of validity.

关键词

Keywords

references

WANG Y T, MO W B, BAI M Y, et al. Lightning interference identification method based on voltage traveling wave waveform characteristics for flexible DC transmission lines [J]. Power system protection and control, 2024, 52(8): 31-44. DOI: 10.19783/j.cnki.pspc.230947.

MENG D, CHEN Z H, XU Y, et al. The impact of extreme weather on the entire process of wind power development and response strategies [J]. Southern energy construction, 2025, 12(2): 1-14. DOI: 10.16516/j.ceec.2024-334.

CHEN R, ZHENG S M, LIU B , et al. Simulation analysis of the interaction of flexible HVDC transmission lines erected on same tower under lightning conditions [J]. Shandong electric power, 2025, 52(7): 35-43. DOI: 10.20097/j.cnki.issn1007-9904.2025.07.004.

ZHANG Y G, GU S Q, LI J, et al. Lightning activity prediction method based on deep learning and its application in transmission line lightning protection [J]. Insulators and surge arresters, 2024(4): 18-28. DOI: 10.16188/j.isa.1003-8337.2024.04.003.

GUO D M, LIN Z M, LIU G, et al. Modeling and analysis of sliding motion of arc root along OGW during lightning strike [J]. Electric power engineering technology, 2024, 43(5): 181-188. DOI: 10.12158/j.2096-3203.2024.05.018.

WANG Z G, RUAN G Q. Simulation and analysis of lightning protection effect and protection range of distribution line arrester based on ATP-EMTP [J]. Insulators and surge arresters, 2020(1): 66-70. DOI: 10.16188/j.isa.1003-8337.2020.01.011.

ZHOU C S. Lightning protection design of offshore floating photovoltaic power station [J]. Southern energy construction, 2024, 11(6): 102-110. DOI: 10.16516/j.ceec.2024.6.11.

ZEPKA G S, SARAIVA A C V, PINTO O, et al. Lightning forecasting using WRF model over EDP distribution companies areas [C]// Anon. 2013 International Symposium on Lightning Protection (XII SIPDA), Belo Horizonte, Brazil, October 7-11, 2013. New York: IEEE, 2013: 255-260. DOI: 10.1109/SIPDA.2013.6729239.

GONG J Q, GUO F X, TIAN K, et al. The lightning potential forecast with WRF model [J]. Journal of the meteorological sciences, 2014, 34(2): 206-212. DOI: 10.3969/2013jms.0007.

XIONG Y J, LIAO X N, YU B, et al. Warning and forecast system of potential thunderstorm based on weather patterns in Beijing [J]. Journal of catastrophology, 2012, 27(2): 67-71, 81. DOI: 10.3969/j.issn.1000-811X.2012.02.014.

JI X L, HU W D, MU J H, et al. Operational system of thunder/lightning potential forecasting based on similar deviation for Ningxia [J]. Meteorological science and technology, 2012, 40(2): 304-309. DOI: 10.19517/j.1671-6345.2012.02.028.

CAI Y X. Research on lightning potential forecast based on the convective parameter stepwise elimination method [J]. Science technology information, 2011(7): 130-131. DOI: 10.16661/j.cnki.1672-3791.2011.07.211.

SHEN H T, LI F, SHI L, et al. Short-term power load forecasting based on reduction of meteorological data dimensionality and hybrid deep learning [J]. Electric power construction, 2024, 45(1): 13-21. DOI: 10.12204/j.issn.1000-7229.2024.01.002.

DENG S, JIANG Q, MAO X Y, et al. Lightning risk identification and assessment for transmission lines based on the safety risk system [J]. Power systems and big data, 2025, 28(09): 21-29. DOI: 10.19317/j.cnki.1008-083x.2025.09.003.

YAO N P. Research on lightning activity characteristics and nowcasting methods based on multi-source data [D]. Nanjing: Nanjing University of Information Science and Technology, 2021. DOI: 10.27248/d.cnki.gnjqc.2021.000345.

ZHANG J, YAN Z H, ZHANG D Q, et al. Design and application research of an interactive fusion system of multi-source data of electric power and meteorology [J]. Power systems and big data, 2025, 28(02): 83-92. DOI: 10.19317/j.cnki.1008-083x.2025.02.010.

GAO B, JI T C, HUANG T, et al. A study on Impulse current dispersion characteristics of the vertical integration grounding of tower pile foundations [J]. Power system and clean energy, 2025, 41(07): 36-43. DOI: 10.26963/j.psce.2025.07005.

ZHANG Z Y. Research on thunder cloud monitoring and lightning nowcasting method based on multi-source data [D]. Wuhan: Huazhong University of Science and Technology, 2022. DOI: 10.27157/d.cnki.ghzku.2022.001381.

GAO W S, ZHANG B W, ZHOU R X, et al. Nowcasting of the thunderstorm trend based on data collected by lightning location system [J]. Power system technology, 2015, 39(2): 523-529. DOI: 10.13335/j.1000-3673.pst.2015.02.035.

HUANG X Y, MA Y R, HU S M. Extrapolation and effect analysis of weather radar echo sequence based on deep learning [J]. Acta meteorologica sinica, 2021, 79(5): 817-827. DOI: 10.11676/qxxb2021.041.

LIN T Y. Research on lightning prediction using deep spatiotemporal neural network models [D]. Beijing: Beijing Jiaotong University, 2020. DOI: 10.26944/d.cnki.gbfju.2020.001282.

LIN M. Research on prediction model of lightning activity based on improved fuzzy neural network [D]. Xi'an: Xi'an University of Electronic Science and Technology, 2014.

MA C L, YAO X Y, SONG M Q, et al. Analysis of the impact of substation grounding network on the resistive coupling of the surrounding gas pipeline system and the newly built grounding system under high frequency lightning currents [J]. Power system protection and control, 2025, 53(9): 166-175. DOI: 10.19783/j.cnki.pspc.240846.

KOU X S, WANG D L, KE J Y, et al. Risk early warning method of severe convective disasters for transmission lines based on radar echo and LSTM [J]. Electric power engineering technology, 2024, 43(5): 189-198. DOI: 10.12158/j.2096-3203.2024.05.019.

WU A K, DING M, LI D, et al. Lightning potential prediction in Guizhou Province based on physical parameters [J]. Insulators and surge arresters, 2023(1): 208-215. DOI: 10.16188/j.isa.1003-8337.2023.01.025.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

A Short-Term Calgorithm Based on Improved LSTM Neural Network

Analysis of the Distribution Characteristics of Mountain Fires Based on the Disaster Data of Hubei Transmission Lines

Simulation and Performance Analysis of a Hybrid TEC-Liquid Cooling Thermal Management System For the Lithium-Ion Battery Pack

Dual-Layer Optimized Configuration of Photovoltaic Storage Charging Station Based on Orderly Charging of Electric Vehicles

Compressed Air Supply System Based on Binary Gas Compression Energy Storage

Related Author

Sheng GAO

Peihua XU

Zhenghong CHEN

Chi CHENG

Limei YE

Junjie HUANG

Zhengxu GAO

Jun WAN

Related Institution

Hubei Provincial Meteorological Service Center

Wuhan Regional Climate Centre

Three Gorges National Climatological Observatory

Key Laboratory of Basin Heavy Rainfall, China Meteorological Administration

State Grid Hubei Electric Power Research Institute

AI问答

⁰