Research on hysteretic performance prediction method of novel connection dampers for external wall panels based on the mining of temporal combination characteristics

Finite element modeling and experimental test are effective methods to establish constitutive models and evaluate the seismic performance of novel dampers designed to connect external wall panels to steel structures. However, optimizing the key structural parameters of these dampers to suit specific structural systems often presents technical challenges, including low execution efficiency and limited repeatability of the constitutive model. To address these limitations, this study proposes a new method for predicting the hysteretic performance of the novel connection dampers based on temporal combination characteristics mining. First, the critical structural parameters influencing the seismic performance of the novel dampers are identified through refined finite element modeling analysis. Secondly, the temporal characteristics dataset of novel connection dampers is constructed, incorporating these parameters and the corresponding loading protocol. Subsequently, by integrating Convolutional Neural Networks (CNN) and Series Decomposition Block (SDB), the mining model of temporal combination characteristics (MTCC) for the novel connection damper is established. And the method for predicting the hysteresis force of dampers based on temporal combination characteristics mining is developed. Finally, the proposed method is validated using the real measurement examples. The results indicate that the average deviation between the seismic performance index calculated by the proposed method and the measured results is less than 5 %, demonstrating the method's high accuracy and reliability.