Applying Autonomous Learning Algorithm to Movement Balance Control on the Robot

In order to solve the movement balance problems about the two-wheeled self-balance robot, an autonomic learning method is presented. This method is based on the fuzzy adaptive algorithm, and it could identify online the fuzzy model of the robot, and detect the parameter variation of the robot and track its characteristics about the parameter variation over time. This paper uses the model of the robot and the expected performance index to design a fuzzy controller, so that the autonomic learning method was formed, and the stability of this algorithm is proved theoretically. The simulation results show that the autonomic learning method could realize the standing balance and speed tracking of the robot, in the case of deviating from a larger angle to the vertical position. It embodies the higher dynamic response and steady accuracy.