An Intelligent Controller Based on Fuzzy Target Acquired by Reinforcement-learning

The cooperation system operated with the human and the computer is important. In this paper, the fuzzy target knowledge that decides the auxiliary target to reach the final goal by the reinforced learning is acquired. An intelligent cooperated control system that decides the control instruction based on this fuzzy target and the driving knowledge is proposed. The human easily understands the response to the change in the situation of this intelligent control system. The simulation result showed that the system used the fuzzy knowledge acquired by this reinforcement learning was able to achieve a flexible driving in cooperation with the human