Použití modifikovaného Q-uČení pro řizení inverzního kyvadla

Using Modified Q-learning With LWR for Inverted Pendulum Control

článek ve sborníku ve WoS nebo Scopus

en

Locally Weighted Learning (LWR) is a class of approximations, based on a local model. In this paper we demonstrate using LWR together with Q-learning for control tasks. Q-learning is the most effective and popular algorithm which belongs to the Reinforcement Learning algorithms group. This algorithm works with rewards and penalties. The most common representation of Q-function is the table. The table must be replaced by suitable approximator if use of continuous states is required. LWR is one of possible approximators. To get the first impression on application of LWR together with modified Q-learning for the control task a simple model of inverted pendulum was created and proposed method was applied on this model.

Lokálně vážená regresse (LWR) patří mezi aproximátory s lokálním modelem. V tomto přispěvku se zabýváme možnostmi jak rozšířit metodu Q-učení o výhody které přináši metoda LWR.

Locally Weighted Learning (LWR) is a class of approximations, based on a local model. In this paper we demonstrate using LWR together with Q-learning for control tasks. Q-learning is the most effective and popular algorithm which belongs to the Reinforcement Learning algorithms group. This algorithm works with rewards and penalties. The most common representation of Q-function is the table. The table must be replaced by suitable approximator if use of continuous states is required. LWR is one of possible approximators. To get the first impression on application of LWR together with modified Q-learning for the control task a simple model of inverted pendulum was created and proposed method was applied on this model.

Q-Learning, LWR, Continuous Space

2003

24.03.2003

Institute of Mechanics of Solids, Brno University of Technology

Brno

80-214-2312-9

Mechatronics, Robotics and Biomachanics 2003

2