Research On Control For Planar Overactuated Parallel Mechanism

Compared with serial mechanism, parallel mechanism has many advantages such as high precision, strong bearing capacity, etc., which can make up for the lack of serial mechanism.However, due to the existence of singularities inside, the real work space is smaller than its theoretical space. Overactuated parallel mechanism can overcome the weaknesses of ordinary parallel mechanism by eliminating the singularity point, optimizing the input power, increasing the error-tolerant rate of the mechanism and so on, and has simple structure, good controllability, significant economy, has becoming a research hotspot which has great important significance and value.This dissertation use the2-DOF planar overactuated parallel mechanism which has four branches as the research object. First, it analyzes the kinematics aspects and create the kinematic model of this overactuated parallel mechanism, obtained the Jacobian velocity matrix, working space, the torque transmission performance and etc.Second, by establishing and analyzing the dynamic model of this studied mechanism, the transform relationship between the speed of the drive space, the speed of torque limit the end actuator space and the acceleration limits was obtained.Then, this dissertation analyzed several common control methods to discuss their advantages and disadvantages and studied the PD control strategies, the extended PD control and the torque control strategy.Finally, through the builded models, simulated in Matlab for the PD controller, computed torque controller and the robust controller.As can be seen, computed torque control and robust control can eliminate at the error of the end of the mechanism trajectory tracking and external interference and have high accuracy. At last, when done the circular trajectory tracking control experiment by using robust control algorithm, it is found that when this dissertation was moving, the end of the mechanism is relatively stable, and the tracking error converges to0which showed that robust control is effective.