This research project aims at developing modern artificial intelligence-based methods for condition monitoring of electromechanical energy conversion systems, or powertrains. To ensure safe and efficient operation of these powertrains, it is essential to predict their incipient faulty operations at an early stage. By combining experimental results on hardware with simulation results, we will produce synthetic augmented data to be used to train the artificial intelligence (AI) algorithms. These algorithms will also combine data from different application domains, allowing the transfer learning. Moreover, AI will guide the simulation setups to optimally invest the computational resources into relevant simulations. The results of the project are expected to produce new knowledge on how to optimally leverage AI algorithms for energy conversion systems. We will also build a variety of simulation models, which can be used for other purposes such as system optimization and control design.

2020

2024