Policy gradient methods for robot control

摘要

Abstract. Reinforcement learning offers the most general framework to take traditional robotics towards true autonomy and versatility. However, applying reinforcement learning to high dimensional movement systems like humanoid robots remains an unsolved problem. In this paper, we discuss different approaches of reinforcement learning in terms of their applicability in humanoid robotics. Methods can be coarsely classified into three different categories, i.e., greedy methods, ‘vanilla ’ policy gradient methods, and natural gradient methods. We discuss that greedy methods are not likely to scale into the domain humanoid robotics as they are problematic when used with function approximation. ‘Vanilla’ policy gradient methods on the other hand have been successfully applied on a real-world robot. We demonstrate that these methods can be significantly improved using the natural policy gradient instead of the regular policy gradient. Proofs are provided that Kakade’s average of the natural gradient [10] is indeed the true natural gradient. A general algorithm for estimating the natural gradient, the Natural Actor-Critic algorithm, is introduced. This algorithm converges with probability one to the nearest local minimum in Riemannian space of the cost function. The algorithm outperforms non-natural policy gradients by far in a cart-pole balancing evaluation, and offers a promising route for the development of reinforcement learning for truly high-dimensionally continuous state-action systems. 1