Neural network based reinforcement learning control of autonomous underwater vehicles with control input saturation | IEEE Conference Publication | IEEE Xplore