Mobile robot control based on hybrid neuro-fuzzy value gradient reinforcement learning | IEEE Conference Publication | IEEE Xplore