Driving Reliability with Machine Learning and Improving Operation by Digitalization of Power Transformers

For the new challenges of energy supply a new generation of analytical capabilities with the ability to provide deeper insights is mandatory. Machine learning in combination with artificial intelligence can become the new methodological approach for the diagnostics of equipment in the distribution network. In that way machine learning is based on pattern and evaluation of statistical values and herewith it generates knowledge from experience. The decision-making process is often impaired by unplanned outages. In order to efficiently operate and maintain existing systems and prevent critical failures on power transformers, the right monitoring and control system must be present. This paper introduces also an advanced method for analysis, visualization and interpretation of vibroacoustic of online measurements of on-load tap-changers, particularly in relation to their application with transformers in service.