- 期刊
Review of Deep Reinforcement Learning for Real Robots
摘要
Deep reinforcement learning is one of the most exciting fields in artificial intelligence, combining reinforcement learning with the power of deep neural networks to understand the world and act on that understanding. In the past few years, deep reinforcement learning has been extensively studied, with remarkable progress and widespread success in different fields. For robot control, deep reinforcement learning algorithms hold the promise of achieving human‐like tasks or surpassing them. This paper reviews the research status of reinforcement learning algorithms in the field of robot control. The basic theory of reinforcement learning, the mathematical background, and the problem of narrowing down current robotics applications are also included in this review. Finally, future research directions for reinforcement learning are discussed.