Model-based reinforcement learning for infinite-horizon approximate optimal tracking | IEEE Conference Publication | IEEE Xplore