Research On Flexible Body Dynamics Modeling Analysis And Intelligent Control Method Of Lightweight Parallel Robot

Lightweight parallel robots are widely used in circuit board processing,precision sorting,transport and packing of food industry and other fields.However,the mechanical structure and dynamics characteristics of lightweight parallel robots are complex,under the high speed and complex working condition,the position accuracy,working efficiency and operation stability of end-effector will be reduced by the influencing factors such as flexible link,flexible joint and joint clearance,the performances of existing robot control methods are unsatisfactory,there are many problems,such as great error,complex algorithm structure,poor robustness and portability.To solve these problems,this dissertation deeply analyzed the dynamics modeling process and characteristics of lightweight parallel robot,and systematically studied the intelligent control method,identified the variables,the relationship between variables and which variables were controllable in the process of achieving the desired motion state,guided the formulation of intelligent control strategy,to realize the precise,efficient and stable control of the end-effector of lightweight parallel robot.This dissertation took Delta parallel robot as the object of verification,and the specific research work was as follows:(1)Based on Hamilton motion equation and Lagrange principle,a parallel recursive dynamics modeling method was proposed for lightweight parallel robot.The dynamics model of flexible link was established by virtual joint velocity,constraint impulse and canonical momentum derivative.Based on the generalized force and energy of flexible joint,the dynamics model of flexible joint could be deduced.Through analyzing the kinematics constraints and dynamics constraints of flexible bar-groups,the comprehensive dynamics model of lightweight parallel robot was derived.(2)The dynamics coupling effect index was defined by principal transformation.Taking the coupling effect index as objective function,and planning trajectory,which could be used to reduce the process vibration error and residual vibration error of end-effector while keeping mechanical structure unchanged,and the relationship between the coupling effect index and the comprehensive position error of end-effector was also analyzed.(3)Referring to the concept of workspace density and based on comprehensive dynamic dexterity,the workspace latticization method of lightweight parallel robot was further studied.This method could control the trajectory error by adjusting the lattice size according to the requirements of accuracy and efficiency,and the relationship between accuracy and efficiency could be balanced flexibly.(4)The position error chaotic phenomenon of end-effector was identified by Poincaré map and Lyapunov exponent.And based on the concepts of information transfer model and entropy,the chaotic information transfer model and chaotic entropy were deduced,and the chaos suppression system was designed.(5)Based on the dynamics characteristics of lightweight parallel robot,the fuzzy system-fuzzy neural network-backstepping control(FS-FNN-BSC)algorithm was studied,the dynamics coupling effect index was used to plan the trajectory and the workspace was latticed,which could generate the input signal of FS-FNN-BSC,and this study combined with the chaos suppression system to build the intelligent control scheme of lightweight parallel robot,and carried out the detailed experimental demonstration.