Bibliographic citations
Fernández, D., (2024). Desarrollo de una red neuronal artificial para identificar la curva característica de un sensor de potencial de hidrógeno PH en una empresa de monitoreo de la calidad del agua [Tesis, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/27521
Fernández, D., Desarrollo de una red neuronal artificial para identificar la curva característica de un sensor de potencial de hidrógeno PH en una empresa de monitoreo de la calidad del agua [Tesis]. PE: Universidad Nacional de Ingeniería; 2024. http://hdl.handle.net/20.500.14076/27521
@misc{renati/960960,
title = "Desarrollo de una red neuronal artificial para identificar la curva característica de un sensor de potencial de hidrógeno PH en una empresa de monitoreo de la calidad del agua",
author = "Fernández Villanueva, David Benjamín",
publisher = "Universidad Nacional de Ingeniería",
year = "2024"
}
In this work, the characteristic curve of a pH sensor is satisfactorily identified through the training and validation of an Artificial Neural Network. This statement results from the acceptance of the general hypothesis "The Development of an Artificial Neural Network allows the identification of the Characteristic Curve of a Hydrogen PH Potential Sensor in a water quality monitoring company." This was achieved using an experimental methodology, in which a perceptron architecture was used with 6 inputs, 35 neurons in the hidden layer, and 1 neuron in the output layer, the weights were initialized using the Widrow and Nguyen method and for training The general Bayesian regularization algorithm was used. Each of the specific hypotheses was accepted due to the optimal results of Chapter V "Analysis of results" with respect to the sum of squared errors, effective number of parameters, linear regression of the objectives and outputs of the ANN, correlation coefficient, histogram of errors, the comparison of the measurements of the pH degree of three different substances using the electronic module developed with the trained ANN and an electronic measuring device (factory) were similar values with a maximum error of 0.057pH. We conclude this work by making recommendations as a starting point to expand this work.
This item is licensed under a Creative Commons License