Fundamental period prediction of infill reinforced concrete structures using an ensemble of regressors

作     者：Vijayan, Vidya Joy, Chinsu Mereena Shailesh, S. 

作者机构：Department of Civil Engineering TKM College of Engineering Kollam India Department of Computer Science CUSAT Kochi India 

出 版 物：《Asian Journal of Civil Engineering》 (Asian J. Civ. Eng.)

年 卷 期：2024年第25卷第7期

页      面：5559-5570页

主　　题：Ensemble regression Fundamental period Infill walls Reinforced concrete structures 

摘      要：The fundamental period plays an important role when a structure is designed for seismic load. Infill walls are non-load-bearing walls created mostly from masonry, concrete, and other heavy materials, filled in the primary structural frame for a proper structural cladding system. As a result, this infill wall will increase the stiffness of the structure, thereby fundamental time period is significantly changed. Most of the studies on the fundamental period do not give much importance to the infill walls even though it is crucial to be analyzed. In this work, we propose an automated and efficient analysis method for predicting the fundamental period of infill Reinforced Concrete frames using machine learning techniques. As the nature of dependency of different independent variables considered in this study is unknown, different regression techniques were chosen for this purpose. So, we rely upon an exceptional machine learning technique called ensemble learning, which combines predictions from different models to deduce the final prediction more accurately. The storey numbers, the number of spans, length of span, stiffness of infill wall, and percentage of openings are set as input factors, while the value of the fundamental time period is chosen as an output. The proposed regression model s correctness is verified by comparing it to existing formulae in the literature. As a result, in comparison to statistical models, the linear regression model shows an r2 value of 0.98921 and has better ability, flexibility, and accuracy. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2024.