Título :	Compresión y minería de datos para el diagnóstico de fallas incipientes de cortocircuito en transformadores monofásicos con carga eléctrica
Sustentante:	Arantxa Contreras Valdes
Palabras clave :	INGENIERÍA Y TECNOLOGÍA CIENCIAS TECNOLÓGICAS INGENIERÍA Y TECNOLOGÍA ELÉCTRICAS
Fecha de publicación :	28-sep-2020
metadata.dc.degree.department:	Facultad de Ingeniería
metadata.dc.degree.name:	Maestría en Ciencias (Mecatrónica)
Descripción :	Timely detection of short-circuit faults in transformers is an important challenge to ensure reliability and safety in the generation and transmission of electrical energy. With this detection, the reduction of costs and time required for device repair can be achieved, and at the same time be able to avoid severe damage to the equipment and large economic losses for both electricity providers and end users. Investigations related to the detection of electrical faults in transformers have shown that the most common fault that occurs in these is the short-circuit fault. The methodology developed in this research work allows the diagnosis of the incipient fault condition of a single phase transformer with electric load. In general, the methodology is based on the acquisition of multiple physical quantities such as temperature, current, voltage and vibrations. In this sense, it should be taken into account that the methodology presented proposes the implementation of a data compression algorithm based on a discrete wavelet transform to increase performance during data characterization and eliminate the noise contained in the signals. Then, the data is decompressed for analysis. The variance is extracted and subsequently the results are standardized to be analyzed by a data mining algorithm. A support vector machine is used as a classifier in order to detect the severity of the fault under different load conditions. The results obtained reflect the effectiveness and usefulness of the proposed methodology since it is possible to identify and classify the short-circuit faults induced in the transformer.
URI :	http://ri-ng.uaq.mx/handle/123456789/2349
Otros identificadores :	Incipient fault detection Transformer Data compression Data mining Support vector machine
Aparece en las colecciones:	Maestría en Ciencias (Mecatrónica)

Los ítems de DSpace están protegidos por copyright, con todos los derechos reservados, a menos que se indique lo contrario.