Desarrollo de un código de programación para el mantenimiento preventivo de transformadores de potencia
Power transformers are the most important and, therefore, costly equipment inside a power substation, whether it is the elevation-power in a generating plant, transmission or sub-transmission in pass-through substations, or reduction-power for both public marketing companies and private companies th...
| Author: | |
|---|---|
| Format: | master thesis |
| Status: | Published version |
| Publication Date: | 2022 |
| Country: | Ecuador |
| Institution: | Universidad Técnica de Cotopaxi |
| Repository: | Repositorio Universidad Técnica de Cotopaxi |
| Language: | Spanish |
| OAI Identifier: | oai:oai:repositorio.utc.edu.ec:27000:27000/9790 |
| Online Access: | http://repositorio.utc.edu.ec/handle/27000/9790 |
| Access Level: | Open access |
| Keyword: | TRANSFORMADOR SVM ALGORITMOS DIAGNOSTICO ELECTRICIDAD |
| Summary: | Power transformers are the most important and, therefore, costly equipment inside a power substation, whether it is the elevation-power in a generating plant, transmission or sub-transmission in pass-through substations, or reduction-power for both public marketing companies and private companies that feed their production processes, according to these different types of uses, the equipment is subject to different environmental conditions such as degree of contamination, salinity, height, and relative humidity, among others, as well as factors of demand and quality of energy, which causes different electrical efforts and thermal situations inside the transformer. Consequently, it is necessary to implement different diagnostic techniques according to the stages of the maintenance plan, depending on the conditions. Hence, the following project aims to execute a programming code based on the Support Vector Machine (SVM) supervised learning algorithm o process the concentrations of gases in the mineral dielectric oil obtained from the Dissolved Gas Analysis (DGA). To interpret the data, the Duval method exposed in the IEEE C57.104™-2019 standard is used in such a way that form that serves for verification and comparison with standardized and nonstandardized methods such as Artificial Neural Networks (ANN), contributing to the decisions made in the preventive maintenance program to be implemented in the future in the different power transformers. |
|---|