竞争式Takagi-Sugeno模糊再励学习

COMPETITIVE TAKAGI-SUGENO FUZZY REINFORCEMENT LEARNING

针对连续空间的复杂学习任务 ,提出了一种竞争式 Takagi- Sugeno模糊再励学习网络(CTSFRLN) ,该网络结构集成了 Takagi- Sugeno模糊推理系统和基于动作的评价值函数的再励学习方法 .文中相应提出了两种学习算法 ,即竞争式 Takagi- Sugeno模糊 Q-学习算法和竞争式 Takagi- Sugeno模糊优胜学习算法 ,其把 CTSFRLN训练成为一种所谓的 Takagi- Sugeno模糊变结构控制器 .以二级倒立摆控制系统为例 ,仿真研究表明所提出的学习算法在性能上优于其它的再励学习算法 .

This paper proposes a competitive Takagi-Sugeno fuzzy reinforcement learning network (CTSFRLN) for solving complicated learning tasks of continuous domains. The proposed CTSFRLN is constructed by combining Takagi-Sugeno type fuzzy inference systems with action-value-based reinforcement learning methods. Two competitive learning algorithms are derived, including the competitive Takagi-Sugeno fuzzy Q-learning and the competitive Takagi-Sugeno fuzzy advantage learning. These learning methods lead to so called Takagi-Sugeno fuzzy variable structure controllers. Simulation experiments on the double inverted pendulum system demonstrate the superiority of these learning methods.