Controlling Mixed-Mode Fatigue Crack Growth using Deep Reinforcement Learning