Monitoreo de procesos de conversión de energía mediante métodos de aprendizaje estadístico.
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
En el presente documento se expone el trabajo realizado por el alumno Marcel Ignacio Gutiérrez Gutiérrez para su Memoria de Titulo de nombre “Monitoreo de procesos de conversión de energía mediante métodos de aprendizaje estadístico”, en donde el alumno presenta una serie de modelos realizados en lenguaje Python (PCA, PLS, KPCA y KPLS) aplicados sobre un contexto real (Paneles Fotovoltaicos) con propósitos comparativos. Durante el documento, el alumno explica las características del dataset utilizado como el procesamiento correspondiente para su uso, el funcionamiento general de los modelos y las métricas utilizadas, en conjunto a la estructura general de los códigos implementados, finalmente seguido de pruebas en donde se aplica anomalías artificiales sobre el dataset (Drift, Noise, Packet Loss, Retained Values y Packet Drop) para probar la eficacia de los modelos con respecto a la detección de fallas.
In the following document it’s presented the work carried out by the student Marcel Ignacio Gutiérrez Gutiérrez for his college degree which it’s named “Monitoring of energy conversion processes through the use of statistical learning methods”. In this document, the student presents a series of models written in Python language (PCA, PLS, KPCA and KPLS) which has been applied on a real case (Photovoltaic Panels, most commonly known as solar panels) in order to compare their perfomances. During the document, the student explains the dataset and it’s features as well as the preprocessing made on the dataset, it also explains the general functioning of the implemented methods together with the metrics utilized for this work, in addition to the structure of the codes produced. Lastly, the student presents the results obtained by the models and the tests performed using artificial anomalies (Drift, Noise, Packet Loss, Retained Values y Packet Drop) in order to check the efficacy and perfomance of the models regarding their ability for fault detection.
In the following document it’s presented the work carried out by the student Marcel Ignacio Gutiérrez Gutiérrez for his college degree which it’s named “Monitoring of energy conversion processes through the use of statistical learning methods”. In this document, the student presents a series of models written in Python language (PCA, PLS, KPCA and KPLS) which has been applied on a real case (Photovoltaic Panels, most commonly known as solar panels) in order to compare their perfomances. During the document, the student explains the dataset and it’s features as well as the preprocessing made on the dataset, it also explains the general functioning of the implemented methods together with the metrics utilized for this work, in addition to the structure of the codes produced. Lastly, the student presents the results obtained by the models and the tests performed using artificial anomalies (Drift, Noise, Packet Loss, Retained Values y Packet Drop) in order to check the efficacy and perfomance of the models regarding their ability for fault detection.
Description
Tesis presentada para optar al título de Ingeniero/a Civil Informático/a.
Keywords
Machine learning, Procesamiento de datos, Interpretación estadística de datos