Text this: Supervised optimal control in complex continuous systems with trajectory imitation and reinforcement learning