Displacement Monitoring Method of Transmission Tower Foot Based on FEM and Deep Learning

Abstract

Transmission tower footing displacement can have a large impact on the stable operation of transmission lines. Traditional structural health monitoring methods have limited accuracy. Application of deep learning methods brings new monitoring solutions for online monitoring of power transmission towers. The amount of tower fault data is limited, which cannot provide a good database for deep learning. In this study, a tower displacement monitoring scheme combining finite element method (FEM) and deep learning (DL) was developed. The response data of wind-induced transmission tower vibration under different fault conditions were simulated by FEM. The tower fault database is expanded by means of data enhancement. The dynamic response tests under different displacement states were carried out, and the displacement monitoring of the tower foot was carried out by deep learning method. The findings of this research program show that it is possible to accurately identify the displacement status of the footing of transmission towers and reduce tower accidents.