Maximilian Schenke, a research assistant in the Department of Power Electronics and Electric Drive Systems (LEA), successfully defended his doctoral dissertation on Monday 18 May 2026. In his thesis, entitled ‘Direct Torque Control of Permanent Magnet Synchronous Motor Drives via Safe Reinforcement Learning’, he demonstrated that artificial intelligence, using reinforcement learning, is capable of controlling an electric drive without prior knowledge. Instead of the typical controller parameterisation based on the drive’s electrical characteristics, the control programme was experimentally learnt on a running test bench within just ten minutes of training, with particular emphasis placed on the safety of the learning phase. In the field of electric drive control, Mr Schenke is thus one of the first to have successfully applied reinforcement learning.
Maximilian Schenke successfully defends his doctoral dissertation
