Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles

Several empirical and theoretical methods have been used in civil infrastructure to ascertain the deep foundation’s load capacity. The models in this scenario are primarily driven by physical presumptions as well as the construction of estimations utilizing mathematical frameworks. In this article,...

Full description

Saved in:
Bibliographic Details
Main Authors: Yan Peng, Haiquan Gao
Format: Article
Language:English
Published: Tamkang University Press 2025-06-01
Series:Journal of Applied Science and Engineering
Subjects:
Online Access:http://jase.tku.edu.tw/articles/jase-202602-29-02-0023
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850122110508204032
author Yan Peng
Haiquan Gao
author_facet Yan Peng
Haiquan Gao
author_sort Yan Peng
collection DOAJ
description Several empirical and theoretical methods have been used in civil infrastructure to ascertain the deep foundation’s load capacity. The models in this scenario are primarily driven by physical presumptions as well as the construction of estimations utilizing mathematical frameworks. In this article, innovative design patterns were developed, and three hybrid adaptive neuro-fuzzy inference systems (ANFIS) optimized with artificial rabbit optimization (ARO), cuckoo optimization algorithm (COA), and grey wolf optimization (GWO) have been applied to use experimental data to calculate the driven piles’ bearing capacity (Qt). To increase the optimal networks’ modeling efficacy, optimization methods were deployed to determine the essential parameters of the simulations. Also, other algorithms were developed for comparison purposes, such as single ANFIS, support vector regression (SVR) M5P, multi-adaptive regression spline (MARS), random forests (RF), and random trees (RT). It was concluded that both ANFIS systems optimized with ARO, GWO, and COA accomplish admirably among the categories of trains and tests, with a minimum R^2 of 0.9285 in the learning dataset and 0.9313 in the examining dataset, respectively, indicating a strong similarity between experimental and estimated Qt. Comparing the outcomes of the single and hybrid models, the highest performance belonged to ARO-ANFIS, by gaining the largest values of correlation metrics and the lowest values of error-based metrics. After examining the dependability and considering the justifications, the ANFIS paired with ARO outperformed the COA-ANFIS and GWOANFIS in the Qt of driven piles forecasting model, this is the suggested system.
format Article
id doaj-art-25317fe2a7c0409d857043a5cec5c367
institution OA Journals
issn 2708-9967
language English
publishDate 2025-06-01
publisher Tamkang University Press
record_format Article
series Journal of Applied Science and Engineering
spelling doaj-art-25317fe2a7c0409d857043a5cec5c3672025-08-20T02:34:55ZengTamkang University PressJournal of Applied Science and Engineering2708-99672025-06-0129248349810.6180/jase.202602_29(2).0023Hybrid ANFIS systems: Evaluation of bearing capacity of driven pilesYan Peng0Haiquan Gao1Anhui Technical College of Industry and Economy, Hefei, Anhui, 230051, ChinaCSCEC4 CIVIL ENGINEERING CO.LTD, Hefei, Anhui, 230000, ChinaSeveral empirical and theoretical methods have been used in civil infrastructure to ascertain the deep foundation’s load capacity. The models in this scenario are primarily driven by physical presumptions as well as the construction of estimations utilizing mathematical frameworks. In this article, innovative design patterns were developed, and three hybrid adaptive neuro-fuzzy inference systems (ANFIS) optimized with artificial rabbit optimization (ARO), cuckoo optimization algorithm (COA), and grey wolf optimization (GWO) have been applied to use experimental data to calculate the driven piles’ bearing capacity (Qt). To increase the optimal networks’ modeling efficacy, optimization methods were deployed to determine the essential parameters of the simulations. Also, other algorithms were developed for comparison purposes, such as single ANFIS, support vector regression (SVR) M5P, multi-adaptive regression spline (MARS), random forests (RF), and random trees (RT). It was concluded that both ANFIS systems optimized with ARO, GWO, and COA accomplish admirably among the categories of trains and tests, with a minimum R^2 of 0.9285 in the learning dataset and 0.9313 in the examining dataset, respectively, indicating a strong similarity between experimental and estimated Qt. Comparing the outcomes of the single and hybrid models, the highest performance belonged to ARO-ANFIS, by gaining the largest values of correlation metrics and the lowest values of error-based metrics. After examining the dependability and considering the justifications, the ANFIS paired with ARO outperformed the COA-ANFIS and GWOANFIS in the Qt of driven piles forecasting model, this is the suggested system.http://jase.tku.edu.tw/articles/jase-202602-29-02-0023driven pilesbearing capacityevaluationdeep foundationanfisartificial rabbit optimization
spellingShingle Yan Peng
Haiquan Gao
Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
Journal of Applied Science and Engineering
driven piles
bearing capacity
evaluation
deep foundation
anfis
artificial rabbit optimization
title Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
title_full Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
title_fullStr Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
title_full_unstemmed Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
title_short Hybrid ANFIS systems: Evaluation of bearing capacity of driven piles
title_sort hybrid anfis systems evaluation of bearing capacity of driven piles
topic driven piles
bearing capacity
evaluation
deep foundation
anfis
artificial rabbit optimization
url http://jase.tku.edu.tw/articles/jase-202602-29-02-0023
work_keys_str_mv AT yanpeng hybridanfissystemsevaluationofbearingcapacityofdrivenpiles
AT haiquangao hybridanfissystemsevaluationofbearingcapacityofdrivenpiles