Approximating Helical Pile Pullout Resistance Using Metaheuristic-Enabled Fuzzy Hybrids

Piles have paramount importance for various structural systems in a wide scope of civil and geotechnical engineering works. Accurately predicting the pullout resistance piles is critical long-term resilience infrastructures. In this research, three sophisticated models are employed precisely (Pul) helical piles. Metaheuristic schemes gray wolf optimization (GWO), differential evolution (DE), ant colony (ACO) were deployed tuning an adaptive neuro-fuzzy inference system (ANFIS) mapping Pul behavior from independent factors, namely embedment ratio, density class, ratio shaft base diameter to diameter. Based on results, i.e., Pearson’s correlation coefficient (R = 0.99986 vs. 0.99962 0.99981) root mean square error (RMSE 7.2802 12.1223 8.5777), GWO-ANFIS surpassed DE- ACO-based ensembles training phase. However, smaller errors obtained DE-ANFIS ACO-ANFIS pattern. Overall, results show that all capable both loose dense soils with superior accuracy. Hence, combination ANFIS mentioned metaheuristic algorithms recommended real-world purposes.