Ultimate strain estimation of concrete wrapped by aramid fiber employing coati optimization-based systems

The application of fiber-reinforced polymer ( \(FRP\) $\mathit{FRP}$ ) wraps to increase the load-carrying capacity of reinforced concrete structures has certain advantages, such as corrosion resistance, a light weight, high stiffness and easy adaptability to the shape of the structure. In order to estimate the ultimate strain (ε_cu) of circular columns that are wrapped in aramid fiber-reinforced polymer (AFRP), a variety of machine-learning (ML) techniques were created. The Decision Tree (DT) and the Least Square Support Vector Regression (LSSVR) were both developed as a result of this aim. This research used the Coati optimization algorithm (COA), a metaheuristic optimization approach, in conjunction with DT and LSSVR analysis to arrive at precise values for the decision parameters. During training and evaluation phases, the DT-COA technique showed great reliability with R² values of 0.9842 and 0.9879. The pattern indicates that, in terms of R² values, the LS-COA approach fared better than the DT-COA strategy. The equivalents of the approach are 0.9854 and 0.9888. These values were lower than the ones that DT-COA gained. The \(LS-COA\) $LS-COA$ had the smallest uncertainty index values during the training and evaluation phases, with values of 0.342 and 0.2621, accordingly lower than the ones that \(DT-COA\) $DT-COA$ obtained, which were 0.3555 and 0.2731. Both models are trustworthy and accurate, with LS-COA showing a little edge over the other model, as shown by reasoning and evaluation indications.