この資料をSMS送信: Navigation System in Dynamic Indoor Environment Using Deep Reinforcement Learning with Actor-Critic