Research On Position Control Method Of Underactuated Planar Manipulators

Underactuated robot is a kind of manipulator with one or more passivejoint. The major character is that the dimension of input space is less thandimension of output space. We can see that the constraint of the system isun-integral. It is called second order nonholonomic constraint. The researchon control of underactuated robot systems has been a challenging researchfocus because of the decrease of actuated joint.The problem of position control for underactuated planar robot is studiedin this dissertation, and main contents are outlined below:An adaptive fuzzy sliding-mode control strategy is presented to achieveposition control of underactuated manipulators with locker in passive joints.Defining sliding-mode surface as input variable of fuzzy reduces the numberof fuzzy rules, and then the tuning algorithms are derived in the sense of theLyapunov stability theorem, thus the stability of the system can be guaranteed.This method provides the system with the necessary robustness to performtasks regardless of the modeling errors and external disturbances.Aiming at underactuated manipulator with free swing passive joint, wepropose a CMAC neural network based feedforward compensated and PDcontrol scheme, by tracking appropriated off-line planned trajectory achieveaccurate position control. CMAC learns the inverse dynamics model ofsystem and PD controller adjusts the tracking errors, in this way, the dynamicperformance is improved and the robustness is enhanced.The study on position control of underactuated manipulator in workspacethe presence of obstacles. The method of collision free trajectory planning ofsecond order nonholonomic system is introduced in brief, a sliding modetrajectory tracking control method is designed because of the tracking androbust performance of sliding mode control. Simulation results show theeffectiveness of the proposed control approach.