Dynamics-Invariant Quadrotor Control using Scale-Aware Deep Reinforcement Learning