ISBN-13: 9786204732015 / Angielski / Miękka / 56 str.
This book aims at providing basic introduction about reinforcement learning (RL), application of RL as controller and introducing stability in RL. Q-learning is most widely used RL technique and is explained in detail in this book. In RL, "Curse of Dimensionality" is a major issue and author has used fuzzy inference system to handle this problem resulting in Fuzzy Q learning. Reinforcement learning works on exploitation and exploration policy and hence RL based controller may face stability issue. The main emphasis of this book is to introduce stability in RL based controller using Lyapunov theory. The proposed RL based controllers are simulated on various nonlinear systems including Inverted Pendulum and Robotic Manipulator.