Robust Control in the Worst Case Using Continuous Time Reinforcement Learning | IEEE Conference Publication | IEEE Xplore