Catenary action plays crucial role in resisting the applied vertical load at large deformations stage in reinforced concrete (RC) structures. This paper aims to predict the catenary action capacity of RC beam-column substructures by utilizing the distinctive properties of gene expression programming (GEP). The input parameters selected for the modelling are: double-beam span-to-depth ratio, relative axial restraints stiffness, relative rotational restraints stiffness, bottom and top longitudinal reinforcement ratios, and yield strength of longitudinal rebars. A comprehensive and reliable database was collated from internationally published research articles to develop and verify the model. The GEP-based model was assessed by comparing its performance with regression based model. Various statistical indicators and external validation criteria suggested in literature proved that the model is accurate and possess high prediction and generalization capacity. Sensitivity analysis was carried out to show the contributions of the input parameters, while parametric analysis was performed to show that the proposed model is not merely a combination of the input parameters but can accurately represent the given physical system. The proposed formulation from GEP is found to be simple, robust, and easy to utilize for pre-design purposes.

悬链线作用在抵抗钢筋混凝土（RC）结构大变形阶段所施加的垂直载荷方面起着至关重要的作用。本文旨在通过利用基因表达编程（GEP）的独特特性来预测RC梁柱子结构的悬链作用能力。为建模选择的输入参数为：双梁跨度-深度比，相对轴向约束刚度，相对旋转约束刚度，底部和顶部纵向钢筋比以及纵向钢筋的屈服强度。从国际上发表的研究文章中收集了一个全面而可靠的数据库，以开发和验证该模型。通过比较基于GEP的性能与基于回归的模型来评估其性能。文献中提出的各种统计指标和外部验证标准证明该模型是准确的，具有较高的预测和泛化能力。进行了敏感性分析以显示输入参数的贡献，而进行参数分析以显示所提出的模型不仅是输入参数的组合，而且可以准确地表示给定的物理系统。已发现，GEP的拟议配方简单，坚固且易于用于预设计目的。同时进行了参数分析，表明所提出的模型不仅是输入参数的组合，而且可以准确地表示给定的物理系统。已发现，GEP的拟议配方简单，坚固且易于用于预设计目的。同时进行了参数分析，表明所提出的模型不仅是输入参数的组合，而且可以准确地表示给定的物理系统。已发现，GEP的拟议配方简单，坚固且易于用于预设计目的。