Improved Robustness of Reinforcement Learning Based on Uncertainty and Disturbance Estimator | IEEE Conference Publication | IEEE Xplore