Abstract
In resent years, the research on the autonomous mobile robot has been actively performed and it is expected for the robot to participate in various fields. In addition there are also many researches so that the robot might be able to decide flexibly by applying soft computing method such as neural networks. However, the robot may not realize some behaviors desired by the developer in the complex environment. In the case, the desired robot behaviors are often realized by an excellent human expert.
In this research, by making the robot learn the expert's operation, it is confirmed whether the target can be achieved. We used a radio controlled car on the market as a learning objects. An expert controls a radio controlled car with the motion image obtained by the ceiling camera and operation of figure of eight. At this time, some voltage values obtained by the radio controller is used as teaching signals of CMAC learning algorithm. By this experiment, we confirmed that the running
motion in a figure eight is obtained as the control skill like the human operation.
