Work

Approximate optimal trajectory tracking and dynamic obstacle avoidance for affine system via online learning