The National Science Centre has awarded funding for scientific research to Kamila Jankowska, M.Sc. an employee of the Department of Machines, Drives, and Electrical Measurements, as part of the PRELUDIUM 23 competition. The project supervisor is Mateusz Dybkowski, Ph.D., D.Sc., assoc. Prof. The title of the project, with registration number 619025, is "Application of deep learning methods in detection, classification and compensation of the measuring sensors faults in permanent magnet synchronous motor drive system."

The obtained funding will enable research on an innovative system for detection, classification, and compensation of sensor damage, based solely on deep learning techniques, in a drive system with a permanent magnet synchronous motor. To this end, the project plans to conduct research not only on damage detection and classification systems but also on state variable estimation.

Many solutions described in the literature address machine diagnostics based on measurement sensors, while overlooking the possibility of failures in such components. Appropriate diagnostics of measurement sensors in the early stages of failure, based solely on software methods without the use of additional components, can significantly improve the safety level of electric motor control systems.

An important assumption of the project is also to conduct an analysis of the diagnostic system's operation for motors with different rated powers, both in simulation and experimental studies, in order to achieve a possibly universal solution.