Research On SCARA Robot Dynamic Parameter Identification And Feedforward Control

With the continuous expansion of the application range of industrial robots in industrial sites,while ensuring efficiency in various application fields,the accuracy requirements for robots are gradually increasing.Among them,SCARA robot has the advantages of flexibility and speed,and high repeatability.In industrial occasions such as rapid handling,dispensing,laser welding,and precision assembly,the robot is required to work along a fixed path,and appropriate torque must be applied to the robot joints so that the robot end accurately follows the desired trajectory.The robot system has strong nonlinearity.The traditional industrial control method does not consider the nonlinear factors such as the robot's dynamic characteristics,and is no longer suitable for high-precision trajectory tracking control.Effectively improve the robot's trajectory tracking accuracy,and improve the robot's dynamic response rate.Therefore,it is of great value and significance to identify the parameters in the robot's dynamic model and then use feedforward control to conduct research.First of all,this article takes SCARA robot as the research object,in order to avoid encountering singular positions when designing the trajectory of the robot,the kinematics of the robot model is studied.By establishing the D-H coordinate system,the kinematics forward and inverse solutions of the robot are solved,and then the singular position of the robot is obtained through the forward and inverse kinematics relationship.The SCARA robot experimental platform and specific parameters are introduced to lay the foundation for subsequent experiments.Secondly,to solve the problem of inaccurate parameters in the SCARA robot dynamics model,a method of dynamic parameter identification using recursive least squares is proposed.The robot model is dynamically modeled,and the dynamic equations are simplified and linearized according to actual needs and verified to obtain the parameter set to be identified in the dynamic model.Then,an excitation trajectory optimization scheme is designed,and the regression matrix condition number in the dynamic equation is used as the optimization index.The optimal excitation trajectory of the robot is obtained and verified by experiments.Next,a dynamic parameter identification scheme was designed.After many experiments and data filtering,the recursive least square method was used to identify the robot dynamic parameters,and another trajectory was selected for verification,which proved the correctness of the parameter identification results.Finally,for the problem of insufficient control accuracy of the robot system,based on the dynamic parameter identification results,a torque feedforward control scheme based on the dynamic model is designed.Compensate the friction force in the joint,at the same time calculate the feedforward torque value required by the robot,and build the ADAMS +MATLAB joint simulation platform for simulation verification.Secondly,the robot control system was developed a second time,and the feedforward control was implemented.The simulation and experimental results showed that the torque feedforward control can effectively improve the trajectory tracking accuracy of the robot and improve the dynamic characteristics.