In this webinar, Prof. Vinodh Kumar will discuss a deep reinforcement learning (RL) control framework for a rotary flexible link system. Deep reinforcement learning (DRL) control combines deep neural network with reinforcement learning to solve multi-dimensional control and optimization problems. The session will discuss the potentials of DRL and how it can be used for tracking and vibration suppression of the flexible link system using a continuous action space RL algorithm. The benefits of DRL control over a model-based control will be demonstrated with videos of several experiments.
Vinodh Kumar E received PhD in Electrical Engineering from Anna University in 2014. He received B.E in Electrical and Electronics Engineering from Madras University, and M.E in Control Systems from Anna University in 2004 and 2007, respectively. He joined Vellore Institute of Technology in 2015 and is currently an Associate professor with school of Electrical Engineering, VIT, Vellore. His research interests include optimal control, adaptive control and state estimation.