Research On Modeling And Intelligent Control Of Pendubot

As a typical underactuated robot system, Pendubot has great value in the nonlinear control and robot field. For the nonlinear characteristics in the system, it is important to study the control algorithm, especially fuzzy control as a non-model method.The thesis introduces the development and research on Pendubot in China and other countries, designs a dynamic model based on Lagrange equation, and establishes a simulation model in order to prove the validity of the model based on the motion and energy behavior of the system. The result of the simulation shows the validity of the model.In order to make the control easy, the motion space of the Pendubot is divided into swing-up space and balance space. In the swing-up space, a strategy based on energy pumping is presented. Two Lyapunov functions are designed to ensure that the energy of the system is always increasing and the second pole has a good posture when Pendubot is close to the balance space. Fuzzy controller based on energy pumping is designed for the swing-up space. The simulation verifies the validity of swing-up controller.In the balance space, a TS fuzzy controller including four single-input and single-output fuzzy TS controllers based on single rule module is designed in order to simplify the design of the fuzzy controller, the decision parameters in TS model are identified and the robustness is tested. Compared with the traditional state-feedback control method, the controller has bigger attraction and better robustness.In order to enter the balance space, a swing-up controller is adjusted to get a better posture and a switch controller is designed. The simulation of the whole system control shows the validity of the control methods.