Use of artificial neural networks for predicting settlement of shallow foundations on cohesionless soils / Mohamed A. Shahin (original) (raw)
Related papers
2019
The present study tries to predict the settlement of shallow foundation on granular soil using a mathematical model. The application of feed-forward neural networks with back propagated algorithm is followed for the same. For the development of ANN model, 193 in situ tests data were collected from the literature. The inputs required for the development of model were the foundation pressure, width of footing and the standard penetration number. The predicted settlement using this model was found to compare favourably with the measured settlement. Further the results of sensitivity analysis indicated that the width of foundation has highest impact on the predicted settlement in comparison to other input variables. The present study confirms the ability of ANN models to predict a complex relationship between the nonlinear data as in present case.
Settlement prediction of shallow foundations on granular soils using B-spline neurofuzzy models
Computers and Geotechnics, 2003
The design of shallow foundations on granular soils is generally controlled by settlement rather than bearing capacity. As a consequence, settlement prediction is a major concern and is an essential criterion in the design process of shallow foundations. At present, consistent accurate prediction of settlement of shallow foundations on granular soils has yet to be achieved using many numerical modelling techniques. Recently, multi-layer perceptrons (MLPs) trained with the back-propagation algorithm have been applied successfully to settlement prediction of shallow foundations on granular soils. However, a shortcoming of MLPs is that the knowledge that is acquired during training is distributed across their connection weights in a complex manner that is often difficult to interpret. Consequently, the rules governing the relationships between the network input/output variables are difficult to quantify. One way to overcome this problem is to use neurofuzzy networks in which the acquired knowledge can be translated into a set of fuzzy rules that describe the relationships between the network inputs and the corresponding outputs in a transparent fashion. In the present paper, the ability of neurofuzzy networks to predict settlement of shallow foundations on granular soils and to assist with providing a better understanding regarding the relationships between settlement and the factors affecting settlement is assessed. The sensitivity of neurofuzzy models to a number of stopping criteria is investigated and the models obtained are compared in terms of prediction accuracy, model parsimony and model transparency. The impact of incorporating existing engineering knowledge on neurofuzzy model performance and interpretation is also investigated. The type of neurofuzzy networks used in this research are B-spline networks that are trained with the adaptive spline modelling of observation data (ASMOD) algorithm. The results indicate that B-spline neurofuzzy networks are capable of predicting well the settlement of shallow foundations on granular soils and are able to provide a transparent understanding of the relationships between settlement and the factors affecting it. It is found from this research that neurofuzzy models that use the Bayesian Information Criterion (BIC) are able to strike a balance between model accuracy, parsimony and transparency. The results also indicate that modifying neurofuzzy networks by incorporating existing engineering knowledge can improve model performance and enhance the interpretation of the constructed model. #
Australian Geomechanics, 2002
The problem of estimating the settlement of shallow foundations on granular soils is very complex and not yet entirely understood. The geotechnical literature has included many formulae that are based on several theoretical or experimental methods to obtain an accurate, or near-accurate, prediction of such settlement. However, these methods fail to achieve consistent success in relation to accurate settlement prediction. Recently, artificial neural networks (ANNs) have been used successfully for settlement prediction of shallow foundations on granular soils and have been found to outperform the most commonly-used traditional methods. This paper presents a new hand-calculation design formula for settlement prediction of shallow foundations on granular soils based on a more accurate settlement prediction from an artificial neural network model. The design formula presented is a quick tool from which settlement can be calculated easily without the need for computers.
Deep Pile Foundation Settlement Prediction Using Neurofuzzy Networks
The Open Civil Engineering Journal, 2009
A NeuroFuzzy System (NFS) is one of the most commonly used systems in the real life problems because it has explicit and transparency which results from the fuzzy systems, with the learning and generalization capabilities from the dynamic behavior of the neural networks. It is one of the most successful systems, which introduced to decrement the fuzzy rules that constituting the underlying model. This system has a high efficiency; it gives good results in high speed. The NFS used in this study to predict the settlement of deep pile foundations. The results obtained from this system give good agreement and high precious for prediction of settlement compared with hyperbolic model and statistical regression analysis. Also, this scenario can be applied for similar or more complicated problems in the geotechnical engineering. Keyword: Hyperbolic model, neurofuzzy system, pile foundation, settlement monitoring, statistical analysis. The resultant combined system for the Fuzzy System and the Neural Network is called NFS, which posses the advantage of both, and overcomes some of the drawbacks of individual approaches, such as black-box of neural networks and the limited learning capability of fuzzy systems [3, 5]. One additional advantage of NF networks is that available engineering knowledge can be incorporated into the trained network to optimize model performance and to enhance the interpretation of a constructed model [8].
The Open Civil Engineering Journal
Introduction: Although it is a regular duty of geotechnical engineers to evaluate how much shallow foundation settles in the granular soil, there is no well-approved formula for this task. The intent of this research is to develop a formula that is adequately simple to be used in routine geotechnical engineering work but complete enough to address the behavior of granular soil associated with the settlement issue. Methods: Cone penetration test and foundation load test data were used to generate a formula that can predict the settlement. Genetic Programming (GP) based Symbolic Regression (GP-SR) and artificial neural networks were used to develop an optimized formula. Settlements were also calculated using the finite method and compared to the results of the developed formula. Results and Conclusion: Two formulas were developed using SR, and several models were developed using ANN. ANN model 1 has the highest R2 value (0.93) and the lowest MSE (0.16) among all developed ANN and GP-S...
Applications of Artificial Neural Networks in Foundation Engineering
2000
In recent years, artificial neural networks (ANNs) have emerged as one of the potentially most successful modelling approaches in engineering. In particular, ANNs have been applied to many areas of geotechnical engineering and have demonstrated considerable success. The objective of this paper is to highlight the use of ANNs in foundation engineering. The paper describes ANN techniques and some of
Prediction of settlement of shallow foundations on reinforced soils using neural networks
Geosynthetics International, 2006
The use of reinforcement to increase the bearing capacity and reduce the settlement of shallow foundations is a common construction technique. Although foundation settlement is a major problem for design, few practical methods have been developed to compute the settlement of shallow foundations on reinforced cohesionless soils. In this study, a feedforward backpropagation neural network (BPNN), which is one type of artificial neural network (ANN), is used to predict the settlement of reinforced foundations. The model performance showed very good agreement with the measured settlements. The results indicate that the developed BPNN model may be a powerful tool to accurately predict settlement of shallow foundations on reinforced cohesionless soils.
Journal of Geotechnical and Geoenvironmental Engineering, 2003
Artificial neural networks (ANNs) are a form of artificial intelligence (AI), which in their architecture attempt to simulate the biological structure of the human brain and nervous system. In this report, back-propagation neural networks are used to predict the settlement of shallow foundations on cohesionless soils. More than two hundred cases of actual measured settlements are used to develop and verify the ANN model. The predicted settlements found by utilising ANNs are compared with the values predicted by three commonly used deterministic methods. The results indicate that artificial neural networks are a promising method for predicting settlement of shallow foundations on cohesionless soils, as they outperform the conventional methods.
Soils and Foundations, 2018
Application of the theory of elasticity for the calculation of foundation settlements yields equations that are well-established and consolidated in geotechnical standards and/or recommendations. These equations are corrected by an influence factor to increase precision and encompass the existing complex geotechnical casuistry. The study presented herein utilizes neural networks to improve the determination of the influence factor (Iα), which considers the effect of a finite elastic half-space limited by an inclined bedrock under the foundation. The results obtained with the utilization of artificial neural networks demonstrate a notable improvement in the predicted value of the influence factor in comparison with existing analytical equations.
The application of artificial neural network in geotechnical engineering
IOP Conference Series: Earth and Environmental Science, 2018
The artificial neural network (ANN) is a machine learning technique, which can simulate the physiological structure and mechanism of human brain. The study aims to review the principles of ANN algorithm and their application in geotechnical engineering. Firstly, the basic principles of ANN algorithm are introduced. Secondly, the application of ANN algorithm is presented. The review suggests that ANN can classify soil accurately, and it is able to group rock mass, predict the stability of slopes exactly, which could be used for risk assessment. The predicted settlement that is generated by ANN is near real value.