Monitoring and Prediction of Differential Settlement of Ultra-High Voltage Transmission Towers in Goaf Areas

Critical transmission lines frequently traverse geologically complex mountainous regions, where harsh environments and variable climatic conditions pose significant geohazard risks. Utilizing 163 Sentinel-1A scenes (January 2018 to October 2023), we employed Multi-Temporal InSAR (MT-InSAR) to derive the deformation field along the transmission corridor. Time-series analysis of the Lingshao (LS) line towers, interpreted through the principles of mining subsidence, revealed the mechanisms behind their differential tilt. Simultaneously, time-series deformation at the tower footings was input to a deep learning model for 365-day prediction; the accuracy and practical applicability of which were rigorously assessed. The results demonstrate that (1) a unidirectional subsidence funnel within the transmission corridor deformation field, in the absence of zonal settlement features, strongly indicates the presence of a goaf beneath the line; (2) the integrated approach combining time-series InSAR with the settlement trough method proves feasible for monitoring transmission tower tilt, as validated through field verification; (3) the magnitude and direction of tower tilt correlate directly with their position in the mining-induced subsidence basin, showing convergent tilt in tensile zones, divergent tilt in compressive zones, and uniform settlement in neutral zones; (4) for the eight selected typical tower footings, predicted deformation values ranged from −284.6 mm to −186.3 mm, showing excellent agreement with measurements through correlation coefficients of 0.989–0.999 and Root Mean Square Error (RMSE) values of 0.54–2.17 mm. The framework enables proactive hazard avoidance during line routing and provides early warning for tower defects, significantly enhancing power infrastructure resilience in mining-affected regions.