基础隔震结构基于实时监测数据的多级预警及其在南京博物院老大殿中的应用

The Old Hall of Nanjing Museum is a historical building, which has been uplifted by 3 m and then retrofitted by seismic isolation. Therefore, it is of great engineering significance to monitor the structural safety and establish an early-warning system. A three-level early-warning system was proposed in this paper, and the monitoring data obtained from the Old Hall of Nanjing Museum was taken as research basis to verify the proposed method. Firstly, a dynamic warning threshold design method based on the nonlinear auto-regressive neural network was proposed, which was used to determine the first level threshold. Twenty spectrum-compatible ground motions were selected from the PEER ground motion database, and the second level threshold was taken as the mean displacement of isolation bearings obtained by nonlinear dynamic time history analysis. The third level threshold was set as the specified maximum allowable deformation of isolation bearings under rare earthquakes. Based on the threshold exceeding during a certain period of time, an integrated warning technology was proposed to determine the alarm levels of blue, yellow and red, which served as the basis for the selection of emergency response plan. From the simulation results and monitoring data of the Old Hall of Nanjing Museum, the presented method is proved to be effective in detecting anomalies caused by sudden change of stiffness of isolation bearings.