Research On The Kinematics Calibration And Control Of Spherical 2-DOF Redundant Drive

The workspace of spherical 2-DOF redundant drive parallel mechanism is a spherical surface centered at the intersection of axes of each rotating pair,and there are no singular points in the workspace.It can be used for multi-scene spherical platform,spherical positioning,etc.Therefore,it is required that the mechanism must have high motion accuracy.Based on that background,the kinematics,error model,calibration,dynamics,servo drive system,control strategy,camera calibration and hand-eye calibration of the mechanism are studied,etc.The main research contents are as follows:Firstly,based on the spatial constraints of each member of spherical mechanism,the kinematics inverse solution,velocity and acceleration of spherical 2-DOF redundant drive parallel mechanism are analyzed;the error model of mechanism is established by matrix method and simplified to calibration model,and the calibration process is designed to lay the foundation for calibration experiment;the dynamic model of mechanism is established by Language method,and the driving force of mechanism is solved in SimMechanics.The rigid-body model of mechanism is built in hanics to simulate and verify.Secondly,industrial cameras are installed at the end of the mechanism,and monocular vision "hand-eye" system is introduced to detect the location information of the end of the mechanism;camera imaging geometry model is established based on pinhole model,and camera calibration experiments are carried out using Zhang's calibration method to solve the camera's internal and external parameters;hand-eye calibration experiments using Tsai two-step method are conducted to establish camera coordinate system and mechanism.In addition,in order to improve image quality and filter garbage information,the filtering effects of three image denoising methods,median filtering,mean filtering and Gauss filtering,are compared,and a suitable filtering method is selected to preprocess the above experimental images.Thirdly,the servo drive system of mechanism is studied.In order to realize decoupling control of motor,the mathematical model of permanent magnet synchronous motor is established in d-q coordinate system;space voltage vector control is used instead of sinusoidal pulse width modulation,and space voltage vector switching is used to generate circular rotating magnetic field to provide constant torque for motor;the design,which is based on PI speed controller,sliding mode variable structure speed controller and improved sliding mode variable structure speed controller,is to establish a closed-loop space voltage vector control model of three-phase permanent magnet synchronous motor.The speed tracking effects of the three controllers are compared and analyzed.Finally,an experimental platform is built based on the prototype of spherical mechanism,Beckoff controller and industrial camera;kinematic calibration experiments are carried out on the prototype to solve the geometric errors of the mechanism and correct the structural parameters of the inverse model.After four iterations of experiments,the kinematic accuracy of the end of the mechanism before and after calibration is compared,and the error is reduced to the original one third;the ontology model of the mechanism is built in SimMechanics and the joint simulation model is established as a module in the Simulink with the servo drive system.To further improve the motion performance of the mechanism,a force/position hybrid control strategy based on synchronous coordinator is designed,and the mechanism trajectory tracking simulation and experimental research are carried out by comparing the non-redundant driving control methods.