The construction of new tunnels underneath existing metro lines imposes critical challenges in accurately measuring and forecasting differential settlements, particularly in complex geological and structural settings. This study presents a hybrid measurement and forecasting framework that synergizes sparse strain monitoring with intelligent algorithms to address this issue. The proposed methodology first tackles the inverse problem of full‐line displacement field identification from limited strain gauge data by leveraging a mechanical model derived from the conjugate beam theory. Subsequently, a radial basis function neural network establishes a dynamic correlation between key construction parameters and the identified structural deformations, thereby enabling stage‐wise forecasting of displacement evolution under predefined construction schemes. The proposed framework is validated through controlled numerical simulations and scaled laboratory tests based on the engineering background of Chengdu Metro Line X . Results demonstrate that the method achieves a maximum identification error of 5% for displacement field reconstruction and a maximum prediction error of 7.5% for future displacement trends. The study provides a methodological framework that lays the foundation for proactive risk management in adjacent tunneling projects, enhancing the safety control capabilities under both static and dynamic loading conditions.
Zhang et al. (Thu,) studied this question.