This thesis main topic is under-actuated system, proposing two intelligent control methods and evaluating all these control feasibilities. First of all, hierarchical sliding-mode control method is used to control the second order of under-actuated system ballbot. Then, fuzzy controller is used to adjust hierarchical sliding-mode control parameters, using sliding surface as input, and immediately adjusting sliding surface subsystem's slope and reaching law, also improving dynamic system performance.Right now computer simulation is used to verify the feasibility of controller.