Research On Key Technologies Of Robot Fine Compliance Assembly

Terminal products have the trend of integration currently,which means that it is necessary to assemble all kinds of components in a limited space,that is,assembly accuracy requirements are getting higher and higher.However,in the precision manufacturing of many small devices,there are still many difficulties in the automation of repetitive tasks,such as in the assembly of electronic industry,instrumentation and even toy industry,there are difficulties in the determination of position and orientation of small components,grapping parts,the control of robots and assembly strategy.The assembly of these parts mainly depends on manual work,but manual work can easily lead to low production efficiency,low accuracy and consistency.In order to improve the assembly accuracy of large-scale production lines,special structured environment is usually needed to improve the position accuracy of parts in the production line.For example,the position accuracy of frame on automobile assembly line reaches sub-millimeter level,but this production line is often expensive.Therefore,the fine assembly research is the key to realize the industrial automation production.This paper mainly focuses on the research of parts grabbing,Robot Adaptive control,assembly strategy and compliance strategy in high precision robotic automatic assembly system.Firstly,the structure of Elite EC75 robot is analyzed in this paper,its kinematics is established and validated by connecting rod recurrence method and MATLAB toolbox respectively,Lagrange method is used to model the dynamics of the robot,and the accuracy of dynamics of EC75 is validated with ADAMS.The kinematics and dynamics modeling method is also applicable to the load of the robot.Aiming at the problem of position and posture offset when the tool at the end of the robot clamps the parts,the inertia force,gravity,Coriolis force and centripetal force of the parts are further analyzed under the condition that the traj ectory of the robot,the quality and shape of the parts are known,and the data of the six-dimensional force sensor are used.Consequently,the position and posture deviation of assembly parts are identified online.In the process of robot dynamics modeling,because of the influence of shape and position tolerance,friction between parts,external interference and so on,the nominal value of the dynamic model of the manipulator is inaccurate.In addition,the serial manipulator itself has strong coupling and non-linearity,which makes the control of the manipulator more difficult.In this paper,two kinds of adaptive control methods are used to simulate and validate the dynamics control of the manipulator.The adaptive control based on neural network and robust adaptive PD control are analyzed,and the adaptive rates are given,respectively.Finally,the simulation results and analysis are given by using Simulink.Secondly,aiming at the assembly process of the shaft hole,the analysis and research of the assembly strategy are carried out.Aiming at the problem of jamming in assembly process,this paper proposes a new assembly strategy based on the process of jacking in artificial shaft-hole assembly,and analyses the geometric constraints of the strategy.Compared with the mechanical and physical equation of traditional assembly model,this strategy significantly reduces the computational complexity.Because there is only a single point contact in the assembly process,the analysis process is simple.When there is surface contact in the assembly process due to the elastic deformation of parts,the assembly process can also be effectively analyzed and the unnecessary collision on the assembly body can be reduced.In this paper,a new compliance control method is proposed for the established assembly strategy.Similarly,the flexible strategy in manual assembly is simulated.The model is used to model several teaching data.The parameters of the model are solved by the maximum expectation algorithm.Finally,the control force and path are simulated by the regression strategy of the Gaussian mixture model.Finally,an assembly experiment platform is built according to the assembly strategy and compliance control strategy.According to the specific conditions of the assembly strategy in different stages,the flexibility control effect of the manipulator is verified in stages.Successive servo control experiments,foam slant tracking experiments,aluminum inclined plane tracking experiments,surface tracking experiments and small chamfering assembly experiments are carried out successively.The experimental results verify the correctness of the proposed compliant assembly control strategy.