Adaptive linear quadratic control using policy iteration
Adaptive linear quadratic control using policy iteration
复制标题
DOI：
10.1109/acc.1994.735224
复制
发表时间：
1994-06
期刊：
影响因子：
--
通讯作者：
Steven J. Bradtke;B. Ydstie;A. Barto
中科院分区：
文献类型：
--
作者：
Steven J. Bradtke;B. Ydstie;A. Barto
在本文中，我们提出的稳定性和收敛性的结果，基于动态规划的强化学习应用于线性二次型调节（LQR）。我们分析的具体算法是基于Q-学习，它被证明是收敛到一个最优的控制器，提供的基础系统是可控的，一个特定的信号矢量是持续激励。这是连续问题的基于DP的强化学习算法的第一个收敛结果。
In this paper we present the stability and convergence results for dynamic programming-based reinforcement learning applied to linear quadratic regulation (LQR). The specific algorithm we analyze is based on Q-learning and it is proven to converge to an optimal controller provided that the underlying system is controllable and a particular signal vector is persistently excited. This is the first convergence result for DP-based reinforcement learning algorithms for a continuous problem.