Title: Research on differential lightning protection of 10 kV distribution lines in mountainous area based on classification and recognition of topography

Authors: Yubin Liu; Zhiqiang Xu; Lixing Zhou; Fuzhi Kang; Zeyu Xu

Addresses: State Grid Hunan Electric Power Co., Ltd., Changsha, 410000, China ' State Grid Hunan Electric Power Co., Ltd., Changsha, 410000, China ' School of Electrical and Information Engineering, Changsha University of Technology, Changsha, 410076, China ' State Grid Hunan Electric Power Co., Ltd., Changsha, 410000, China ' School of Electrical and Information Engineering, Changsha University of Technology, Changsha, 410076, China

Abstract: In view of the serious lightning damage of distribution lines in mountainous areas of South China, the relationship between lightning strike and topography of distribution lines is studied. Seven topographical features, such as tower elevation and soil resistivity, are extracted to construct the terrain feature vector representing the lightning probability of tower location, and typical topographical samples of high lightning probability tower are obtained. BP artificial neural network classifier is constructed to classify and identify the topographical features of distribution line tower location in mountainous area. The case application shows that the classifier can correctly identify the lightning risk of tower location. The differential lightning protection measures for distribution lines are proposed, and comprehensive protection measures are taken for lightning prone tower positions. The lightning protection of cross valley section is simulated by using ATPDraw. The results show that the comprehensive lightning protection measures are effective.

Keywords: distribution lines; topography; neural network; differentiated lightning protection.

DOI: 10.1504/IJCIS.2022.128096

International Journal of Critical Infrastructures, 2022 Vol.18 No.4, pp.321 - 335

Received: 06 Aug 2020
Accepted: 30 Oct 2020
Published online: 05 Jan 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article