Design of non-linear controller for unicycle robot based on RBF neural network self-adaption control

In this paper, we focus on the design of non-linear controller for the self-balancing unicycle robot, the robot models according to the Routh equation. On this basis, the RBF neural network self-adaption controller of the unicycle robot is proposed. The real prototype is built. And the validity of the mechanism is verified by experiments, the pitching and the rolling balance could be achieved by the motor driver, the motors are installed on the wheel and the body of the robot. The simulation experiments of the non-linear controller adopts RBF neural network self-adaption control are showed through the Simulink toolbox of Matlab. The simulation results show that the robot could achieve self-balancing after a transient period of time by the designed controller, and the stability and rapidity of RBF neural network control are so prominent that it could satisfy with real-time control.