Original scientific paper
https://doi.org/10.37798/2023722417
Monitoring Transformer Condition with MLP Machine Learning Model
Dino Žanić
orcid.org/0009-0009-0043-9244
; Croatian transmission system operator (HOPS)
Alan Župan
orcid.org/0000-0002-5888-2443
; Croatian transmission system operator (HOPS)
Abstract
Failures of large power transformers in transmission system are always followed by significant costs, which is especially problematic given that they present an unplanned expenditure. In addition from disrupting financial plans, these events can lead to lower system reliability. This paper describes the development and potential application of transformer winding temperature model based on multilayer perceptron class of artificial neural networks. Model is built in Python programming language with data collected over the span of one year for a single transformer. Three input features (oil temperature, winding current and outside temperature) are used in the input layer, aiming to predict the winding temperature in the transformer. By comparing the predicted winding temperature with the actual measured winding temperature, insights into the transformers internal condition can be derived. To demonstrate the models proposed application, two types of transformer condition degradation are simulated and a set of certain indicators based on statistical measures are explored.
Keywords
transformers; artificial neural networks; condition monitoring
Hrčak ID:
306212
URI
Publication date:
15.7.2023.
Visits: 387 *