摘要

In order to achieve a balance between the evaluation accuracy and calculation efficiency in the regional seismic damage prediction method, this paper proposes a regional RC frame structures seismic damage prediction method based on component performance. The nonlinear relationship between the design characteristic parameters, geometric characteristic parameters and seismic damage was constructed by seismic performance evaluation system based on component deformation and machine learning algorithm. The evaluation effect and generalization ability of eight machine learning algorithms were compared, and an efficient prediction method for seismic damage of regional structures was proposed. Finally, a residential area in Guangzhou was taken as an example, the geometric characteristic parameters of regional structures were extracted by UAV technology to realize the whole process of earthquake damage prediction of regional RC frame structures. The result shows that the proposed method can effectively predict the distribution of each component performance state of the regional structures, and intuitively present the performance level of the regional structures. This paper provides an effective method and important application case for seismic damage prediction of regional RC frame structures. ? 2022 Editorial Office of China Civil Engineering Journal.

全文