Steady-State Error Compensation for Reinforcement Learning with Quadratic Rewards | IEEE Conference Publication | IEEE Xplore