مقاله انگلیسی رایگان در مورد پتانسیل مدل غیر پارامتری در پیش بینی ظرفیت باربری فونداسیون – اسپرینگر 2017

 

مشخصات مقاله
انتشار مقاله سال 2017
تعداد صفحات مقاله انگلیسی 7 صفحه
هزینه دانلود مقاله انگلیسی رایگان میباشد.
منتشر شده در نشریه اسپرینگر
نوع مقاله ISI
عنوان انگلیسی مقاله The potential of nonparametric model in foundation bearing capacity prediction
ترجمه عنوان مقاله پتانسیل مدل غیر پارامتری در پیش بینی ظرفیت باربری فونداسیون
فرمت مقاله انگلیسی  PDF
رشته های مرتبط مهندسی عمران
گرایش های مرتبط خاک و پی، سازه
مجله محاسبات عصبی و برنامه های کاربردی – Neural Computing and Applications
دانشگاه College of Education for Human Science/Ibn Rushed – University of Baghdad – Iraq
کلمات کلیدی نزدیکترین همسایه K، محاسبات نرم، رگرسیون خطی چندگانه، ظرفیت تحمل، مدل پیش بینی شده
کلمات کلیدی انگلیسی K-nearest neighbor, Soft computing, Multiple linear regression, Bearing capacity, Predictive model
شناسه دیجیتال – doi
http://doi.org/10.1007/s00521-017-2916-9
کد محصول E8710
وضعیت ترجمه مقاله  ترجمه آماده این مقاله موجود نمیباشد. میتوانید از طریق دکمه پایین سفارش دهید.
دانلود رایگان مقاله دانلود رایگان مقاله انگلیسی
سفارش ترجمه این مقاله سفارش ترجمه این مقاله

 

بخشی از متن مقاله:
1 Introduction

Soil ultimate bearing capacity (qu) is indicated as the least pressure that would cause shear stress failure of the supporting soil immediately below and adjacent to the foundation [1]. The ultimate bearing capacity is also known as the limited settlement of foundation. The importance of this geotechnical term is its major requirement within the foundation design satisfaction. For any geotechnical engineering work (e.g., construction), the allowable qu is controlled by the amount of settlements and their criteria [2]. qu is usually computed via either experimentally or manual calculations through analytical procedure (i.e., empirical formulation). Based on the exist state-of-the-art literature, it has been observed that BC for soil which was proposed by [3] is based on homogenies criteria and on manual calculation method that can be not practically well organized. Several formulas have been proposed by several scholars to compute the BC, such as Terzaghi [3], Meyerhof [4], Hansen [5], and Kumbhojkar [6]. The conceptual theory of all those researches is based on the footing geometry and the shearing resistance angle. The current paper focuses on the utility of the k-nn model as an intelligent approach to predict bearing capacity of soil under shallow strip foundation loading. Soft computing techniques are valuable intelligence models that have been remarkably in various engineering applications in comparison with the classical methods that are difficult to pursue or capture the high complicated relationship between parameters [7–11]. Based on the historical researches conducted in predicting bearing capacity, several studies accomplished the modeling utilization the soft computing techniques [12–16]. In 2005, the most primitive study was conducted in predicting the ultimate bearing capacity using artificial neural network via two main algorithms: backpropagation and radial basis function [17]. The used experimental data belong to shallow foundation on reinforced cohesion-less soil. In summary, the findings approved that the application of ANN outperformed the traditional methods. Another study was conducted for the same application using multiple layer perceptron (MLP) algorithm [18]. The investigated MLP algorithm was verified with the theoretical computation and showed a very comprehended modeling. An adaptive neuro-fuzzy inference system is inspected to capture the correlation between the basic status of soil foundation system and qu [19]. The discovery of the study showed a high potential of the inspected soft computing model to model the bearing capacity. Ornek et al. [20] estimated bearing capacity of circular footing placed on soft clay layer soil. The modeling constructed using artificial neural network (ANN) model. The data set was obtained from an extensive series of field test (i.e., seven different sizing dimensions used as input parameters). The results of ANN model are evaluated with MLR model.