Glass fiber reinforced polymer (GFRP) rebars reinforced in concrete are susceptible to degradation in harsh alkaline environments such as moist reinforced concrete and seawater and sea sand concrete. The residual tensile strength of GFRP rebar is essential in designing guidelines for GFRP reinforced concrete in different codes. The residual tensile strength is reflected as an environment reduction factor (C_E) to incorporate long-term environmental exposure effects. For this purpose, an extensive database comprising 715 tested specimens were collected from literature to develop GEP tree-based model. Aging tests of GFRP rebars were carried out in the laboratory to test the trained model. Initially, nine gene expression programming (GEP) tree-based models were initially developed using RMSE, MAE, and RSE as fitness functions while varying the numbers of genes. The models were developed employing a random selection of 70% of the conditioned specimens for the training purpose in accordance with the literature. The trained models were validated using the remaining 30% data. A model was chosen to create a prediction formula evaluated from the GEP-expression trees (ETs) and derived MATLAB model based on a broader range of statistical errors and correlations. The chosen model was tested using 36 experimental accelerated aging results, which yielded a comparable statistical evaluation to training and validation data. Two types of GFRP rebars, Type-I (volume fraction of 0.50) and Type-II (volume fraction of 0.60) of three different rebar sizes, i.e., 9.5 mm, 12.7 mm, and 15.9 mm were investigated for determining tensile strength retention (TSR) and C_E. The results concluded that smaller Type-I rebars are more susceptible to degradation as compared to Type-II rebars of larger size. A value of 0.76 is recommended for a uniform C_E based on the upper bound of 95% confidence interval for design life of 100 years.

在混凝土中增强的玻璃纤维增​​强聚合物 (GFRP) 钢筋在潮湿的钢筋混凝土和海水和海砂混凝土等恶劣的碱性环境中很容易降解。GFRP 钢筋的残余抗拉强度对于不同规范中 GFRP 钢筋混凝土的设计指南至关重要。残余拉伸强度反映为环境折减系数（C _E) 以纳入长期环境暴露影响。为此，从文献中收集了包含 715 个测试样本的广泛数据库，以开发基于 GEP 树的模型。GFRP 钢筋的老化测试在实验室中进行，以测试训练后的模型。最初，九个基于基因表达编程 (GEP) 树的模型最初是使用 RMSE、MAE 和 RSE 作为适应度函数同时改变基因数量而开发的。根据文献，出于训练目的，采用随机选择 70% 的条件样本来开发模型。使用剩余的 30% 数据验证训练模型。选择一个模型来创建从 GEP 表达式树 (ET) 和基于更广泛的统计误差和相关性的派生 MATLAB 模型评估的预测公式。所选模型使用 36 个实验加速老化结果进行了测试，产生了与训练和验证数据相当的统计评估。研究了三种不同钢筋尺寸（即 9.5 毫米、12.7 毫米和 15.9 毫米）的两种类型的 GFRP 钢筋，即 I 型（体积分数为 0.50）和 II 型（体积分数为 0.60）以确定抗拉强度保留率（ TSR) 和Ç _é。结果得出结论，与较大尺寸的 II 型钢筋相比，较小的 I 型钢筋更容易降解。被推荐用于均匀的0.76的值Ç _é基于上结合95％置信区间为100年的设计寿命。