Reinforcement Learning for Robot Navigation with Adaptive Forward Simulation Time (AFST) in a Semi-Markov Model | IEEE Conference Publication | IEEE Xplore