Research Of Trajectory Planning And Control Strategy Of A3-DOF Hybrid Manipulator

The research contents of this paper consist of dynamic modeling, trajectory planningand trajectory tracking control of a3-DOF serial-parallel hybrid humanoid manipulator.The manipulator is composed of a2-DOF parallel mechanism which forms the elbow anda1-DOF serial mechanism which forms the wrist, and can mimic how a human’s arms areexercised. Serial-parallel hybrid mechanism not only keeps the virtue of parallelmechanism, but also expands the workspace of manipulator. However, the specialstructure of the HHA introduces complexity in dynamic equations, makes trajectoryplanning method special, and places greater demands on control methods. The research oftrajectory planning and trajectory tracking control of the hybrid humanoid manipulator isof great scientific and practical importance.Based on the analysis of the structure of the hybrid manipulator, and taking intoaccount the uncertainties and external disturbance furtherly, the dynamic model of thehybrid manipulator is presented by using Lagrange method.For the sake of making the hybrid manipulator simulating human’s throwing process,the human’s throwing trajectory is obtained by applying the digital image collecting andprocessing technology. To obtain the motion trajectory of the slide tubes of the throwingprocess, the motion of the human’s elbow and wrist is analyzed and then is applied to theelbow and wrist of manipulator. The desired motion trajectory of the three slide tubes ofmanipulator is obtained by using the methods of curve fitting.For the coupling problem of the serial-parallel hybrid manipulator, based on activedisturbance rejection control theory, all the system coupling terms are parts of the totaldisturbance that can be estimated and compensated, finally realizing decoupling control ofthe manipulator. In order to further improve dynamic and stability performance, fuzzycontrollers are designed to adjust the parameters of active disturbance rejection controllertimely. The effectiveness of the designed fuzzy self-tuning disturbance decouplingcontroller is illustrated by simulations.A unique extended state observer based parameter self-tuning sliding-mode controller is designed for the serial-parallel hybrid humanoid manipulator to implement thetrajectory tracking mission. The motivation for using sliding mode control mainly relieson its appreciable features, such as design simplicity and robustness. In order to removethe disturbance impacts of switching gains, the extended state observer is used to estimatethe disturbance in real-time. A parameter self-tuning strategy is proposed to adjust theswitching gain on line according to the running state of the system, greatly eliminating thechatting problem of manipulator, and improveing dynamic performance of the closed-loopsystem. The effectiveness of the ESO-based parameter self-tuning sliding-mode controlleris illustrated by simulations.