The research workplan involves the definition and specification of use cases for the application of AI techniques in solar power plants diagnosis and maintenance, considering the different type of faults, diagnosis needs, available data and the expected performance by end-users for advanced O&M strategies. Then, automatic failure mode detection algorithms will be developed, so that real data can be combined with physical and data-driven models for fault diagnosis, with the possibility of detecting, isolating, and identifying the faults. Aspects such as combination of AI and physical modelling of components (solar inverter) will be the focus of R&D activities, with a view to scientific production oriented to international journals and conferences.

The candidate is required to have a PhD degree in Electrical and Computer Engineering or Power Electronics or related scientific area, and holders of a scientific and professional curriculum showing a relevant profile for the activity to be developed can apply for the tender.

The minimum requirement for the profile is having a Doctorate with reduced post-doctoral experience or without scientific curriculum after the PhD compatible with the category of junior researcher. A minimum of 2 publications in journals or international conferences from the power systems area.

Preference will be given to candidates who have past experience with the power electronics design, simulation and implementation, and domain knowledge about machine learning and data analytics applied to engineering problems.

Scientific advisor: Ricardo Jorge Bessa.