Research on the Protection of Architectural Heritage Based on SBAS and LSTM Technologies

The protection of heritage buildings is of great historical significance. Land subsidence is related to the safety and stability of heritage buildings. Ground monitoring and prediction are the key steps for early warning and timely remediation of heritage buildings. In this paper, remote sensing technology is used to monitor the ground heritage buildings, and neural network LSTM is proposed to predict. Firstly, the SBAS-InSAR technology was used to monitor the land subsidence in the study area in the time series from September 2020 to December 2021. The results show that the maximum settlement rate in the study area reached - 65 mm/yr, the maximum uplift rate reached more than 50 mm/yr, and the maximum cumulative settlement reached - 70 mm. The correlation between the predicted results of the LSTM prediction network constructed in this paper and the actual results reached 0.92, indicating that the LSTM network has accurate prediction performance. Therefore, LSTM and SBAS technology can be used to assist decision-making, early warning and hazard mitigation in the protection of architectural heritage.