Search

Determining technical losses in an electrical system is highly complex due to the large amount of information required for its evaluation. A solution to this problem is the evaluation of losses using an artificial neural network. In this work, a model for evaluating technical losses in subtransmission electrical networks was obtained by using artificial neural networks. This model considers the effective length of the circuit, the maximum apparent and active power, the resistance in the conductors and the number of clients connected to the circuit. The simulation results established a mean square error of 0.0028 and a correlation coefficient between the variables involved of 0.980. The proposed artificial neural network model resulted satisfactory for evaluating technical losses in electrical subtransmission networks. [ABSTRACT FROM AUTHOR]

El artículo investiga la eficiencia de las redes neuronales artificiales (ANN) para la predicción de la capacidad de absorción de energía (g) del concreto de ultra-altas-prestaciones reforzado con fibras (UHPFRC) sometido a tracción directa. Para mejorar el modelo, se dividieron los datos en datos de entrenamiento y testeo. La red se ajustó usando validación k-fold con los datos de entrenamiento y se evaluó con los datos de testeo. El modelo permitió considerar UHPFRC reforzado con una fibra o con mezcla híbrida de dos fibras, de una amplia gama de fibras, tales como fibras de acero rectas, fibras de acero acabadas en gancho, fibras de acero retorcidas, fibras de PVA, fibras de polietileno y fibras de polipropileno. Adicionalmente se realizó una validación experimental de la red. Los resultados demostraron la eficiencia del modelo de acuerdo con los parámetros estadísticos utilizados, así como su precisión y versatilidad para tratar datos nuevos.

This paper investigates the performance of an artificial neural network (ANN) model in predicting the energy absorption capacity (g) of ultra-high-performance fiber reinforced concrete (UHPFRC) under direct tensile test. To avoid overfitting a data division into test and training datasets was carried out. Thereafter the neural networks were trained on the training dataset by using k-fold validation and the result model was evaluated on the test dataset. The model was capable of consider one-fiber or hybrid-two-fibers-blend as reinforced UHPFRC, of a wide range of fibers such as straight steel fibers, hooked end steel fibers, twisted steel fibers, PVA fibers, polyethylene fibers and polypropylene fibers. Experimental works were performed to validate the accuracy of the model on real data. The results demonstrated the efficiency of the model, according to the statistical parameters used for their evaluation, the accuracy and the versatility of the model when new data in considered. [ABSTRACT FROM AUTHOR]

El estudio presentado tiene como objetivo modelar el comportamiento a flexión de los concretos de ultra alto desempeño reforzados con fibras (UHPFRC), incluyendo límite de proporcionalidad (LOP), módulo de rotura (MOR) y sus deflexiones asociadas δLOP y δMOR, utilizando análisis de regresión multivariable y algoritmos de inteligencia artificial (AI). Se construyeron cuatro modelos redes neuronales artificiales (ANN), uno para cada respuesta, con una capa de entrada y una capa oculta, y cuatro modelos de regresión tipo LASSO (least absolute shrinkage and selection operator). Los resultados demostraron la eficiencia de los modelos, evaluados mediante los estadísticos error absoluto medio (MAE), raíz del error cuadrático medio (RMSE), error de sesgo medio normalizado (NMBE) y coeficiente de determinación (R2). Los modelos de redes neuronales mostraron mayor precisión, con valores de R2 de 0.982, 0.969, 0.978 y 0.978, en la predicción de los parámetros (δLOP, LOP, δMOR y MOR) del comportamiento a flexión de los UHPFRC.

This article presents a mathematical model for drying thin layer carrot slices (Daucus carota) using a heat pump dryer (HPD). To select the equation that best describes the drying curve, 10 semi-theoretical and/or empirical models were evaluated. The parameters were determined using the Sigma-Plot® program, and their goodness of fit was compared using the correlation coefficient, R²; Chi-squared, χ²; standard error of the estimate (SEE) and root mean square error (RMSE). Additionally, the effect of the relative moisture, sample thickness and air velocity on the effective diffusivity of the process was evaluated using a response surface tool. Although all the models correctly fit the experimental data, based on the statistical tests, the Wang-Singh model was selected as the best. [ABSTRACT FROM AUTHOR]