Redes neurais artificiais para previsão de capacidade de carga em estacas do tipo hélice contínua / Artificial neural networks for load capacity prediction in continuous flight auger piles

The estimation of the unit weight of soil is carried out using laboratory methods; however, it requires high-quality research material in the form of samples with undisturbed structures, the acquisition of which, especially in the case of organic soils, is extremely difficult, time-consuming and expensive. This paper presents a proposal to use artificial neural networks to estimate the unit weight of local organic soils as leading parameters in the process of checking the load capacity of subsoil, under a direct foundation in drained conditions, in accordance with current standards guidelines. The initial recognition of the subsoil, and the locating of organic soils at the Theological and Pastoral Institute in Rzeszow, was carried out using a mechanical cone penetration test (CPTM), using various interpretation criteria, and then, material for laboratory tests was obtained. The analysis of the usefulness of the artificial intelligence method, in this case, was based on data from laboratory tests. Standard multi-layer backpropagation networks were used to predict the soil unit weight based on two leading variables: the organic content LOIT and the natural water content w. The applied neural model provided reliable prediction results, comparable to the standard regression methods.

Download Full-text

MODELIZACIÓN DE LA RESISTENCIA A LA COMPRESIÓN DEL CONCRETO MEDIANTE REDES NEURONALES ARTIFICIALES

Revista TECNIA ◽

10.21754/tecnia.v23i2.71 ◽

2017 ◽

Vol 23 (2) ◽

pp. 11

Author(s):

Leoncio Luis Acuña Pinaud ◽

Ana Victoria Torre Carrillo ◽

Isabel Moromi Nakata ◽

Pedro Celino Espinoza Haro ◽

Francisco García Fernández

Keyword(s):

Mechanical Properties ◽

Neural Networks ◽

Artificial Neural Networks ◽

Compression Strength ◽

Production Control ◽

Load Capacity ◽

High Capacity ◽

Physical And Mechanical Properties ◽

Structural Element ◽

Artificial Neural

El uso del concreto como elemento estructural va aumentando año tras año. Sin embargo, este producto requiere de unos estrictos controles de calidad sobre sus propiedades mecánicas para el uso como elemento estructural. Este tipo de control implica la existencia de equipos de ensayo con una capacidad de carga de hasta 3.000KN. Sería de gran utilidad para el control de producción la utilización de un método alternativo de gran fiabilidad, que permitiera conocer las propiedades mecánicas a partir de otras propiedades físicas y mecánicas más fáciles de obtener. La alta capacidad de las redes neuronales artificiales (ANN) para modelar los más diversos procesos industriales, las convierte en una herramienta de gran utilidad en el ámbito de la industria del concreto. En este estudio se ha desarrollado una red neuronal para obtener la resistencia a compresión del concreto y se ha modelado dicha propiedad a partir de la composición del concreto y de sus parámetros de fabricación. La red neuronal diseñada, un perceptrón multicapa, ha permitido obtener la resistencia a compresión del concreto con un coeficiente de correlación de 0,97. Esto demuestra la capacidad de las redes neuronales artificiales para obtener la resistencia a compresión del concreto. Palabras clave.-Concreto, Resistencia a compresión, Redes neuronales artificiales. ABSTRACTThe use of concrete as a structural element increases year by year. However, this product needs very stringent control of its mechanical properties in order to be uses as structural element. This type of control requires to have very large testing equipment with a load capacity of up to 3.000KN. Production control would benefit greatly from the use of a highly reliable alternative method that would enable the mechanical properties to be found through more easily obtained physical and mechanical properties. The high capacity of artificial neural networks (ANN) to model a broad range of industrial processes makes them a very useful instrument in the concrete industry. In this study, one neural network was developed to obtain the properties of compressive strength. This property was then modeled though the composition of concrete and manufacturing parameters. The network designed, a multilayer perceptron, allowed the compression strength to be obtained with a regression coefficient of 0,97. This demonstrates the effectiveness of ANN for obtaining the mechanical properties of compression strength of concrete. Keywords.-Concrete, Compression strength, Artificial neural networks.

Download Full-text

Artificial neural network model for predicting load capacity of driven piles

Research Society and Development ◽

10.33448/rsd-v10i1.11526 ◽

2021 ◽

Vol 10 (1) ◽

pp. e12210111526

Author(s):

Alcineide Dutra Pessoa ◽

Gean Carlos Lopes de Sousa ◽

Rodrigo da Cruz de Araujo ◽

Gérson Jacques Miranda dos Anjos

Keyword(s):

Neural Network ◽

Neural Networks ◽

Artificial Neural Network ◽

Artificial Neural Networks ◽

Mean Squared Error ◽

Load Capacity ◽

Pearson Correlation ◽

Driven Piles ◽

Load Tests ◽

Artificial Neural

In geotechnics, several models, empirical or not, have been proposed for the calculation of load capacity in deep foundations. These models run mainly through physical assumptions and construction of approximations by mathematical models. Artificial Neural Networks (ANN), in addition to other applications, are excellent computational mechanisms that, based on biological neural learning, can perform predictions and approximations of functions. In this work, an application of artificial neural networks is presented. The objective here is to propose a mathematical model based on artificial intelligence focused on Artificial Neural Network (ANN) learning capable of predicting the load capacity for driven piles. The results obtained through the neural network were compared with actual values of load capacities obtained in the field through load tests. For this quantitative comparison, the following metrics have been chosen: Pearson correlation coefficient and mean squared error. The database used to carry out the project consisted of 233 load tests, carried out in diverse cities and different countries, for which load capacity, hammer weight, hammer drop height, pile length, pile diameter and pile penetration per blow values were available. These values have been used as input values in a multilayer perceptron neural network to estimate the load capacities of the respective piles. It has been found that the proposed neural model presented, in general, correlation with field values above 90%, reaching 96% in the best result.

Download Full-text