Fault Diagnosis of Power Transformers Based on Sequence-to-Sequence Deep Learning Approaches

The current internship is an extension of an ongoing project (application ref. IT24703) which implies the advanced deep learning-based approaches to diagnose the simultaneous power transformers’ faults based on a time-series dataset.
The power transformers were selected as the candidate assets in the previous project. They are pieces of essential equipment for electricity transmission and distribution systems. The power transformers are expensive and account for massive capital expenditure in electrical networks. Additionally, the reliability and availability of the entire electricity grid depend on its operational stability. So, utility companies must prioritize failure prevention and the sustenance of the optimal functional status of these assets. In this case, detecting the transformer’s incipient faults and the ability to extract insights or useful knowledge from the power transformers’ data in a timely and intelligent way is urgently needed.
Artificial intelligence (AI), particularly machine learning (ML), has grown rapidly in recent years in the context of data analysis and computing that typically allows the applications to function intelligently.

Faculty Supervisor:

Mustapha Nour El Fath;Adnène Hajji

Student:

Partner:

GE Renewable Energy

Discipline:

Engineering

Sector:

Manufacturing; Other services (except public administration); Utilities

University:

Université Laval

Program: