基于ANFIS的单桩竖向极限承载力预测模型

Predicting model of vertical ultimate bearing capacity of single pile using adaptive neuro-fuzzy inference system

人工神经网络已应用在岩土工程的各个方面。针对常用的 BP 网络的不足之处,建立了基于自适应神经模糊推理系统(ANFIS)的单桩竖向极限承载力预测模型。利用文献中桩的载荷试验数据来训练 ANFIS 网络,确定了网络参数。研究结果表明。同常用的 BP 网络相比,ANFIS 预测模型具有学习速度快。拟合能力较好,训练结果唯一等优点,该方法是一种有效地预测单桩极限承载力的方法。

Artificial neural networks have been used in many areas in geotechnical engineering applications.Adaptive neuro-fuzzy inference system (ANFIS) is used to predict vertical ultimate bearing capacity of single pile.The data of pile load test obtained from a literature are used to train ANFIS network and to determine the network parameters.The results show that the proposed modeling approach outperforms classical back-propagation (BP) neural network in terms of computational speed,forecast errors,efficiency. ANFIS is an effective method to achieve both high accuracy and less computational complexity for predicting vertical ultimate bearing capacity.