Research On Visual Servo Control Based On Dynamics Model Of Manipulator

Achieving robot intelligence is an important subject in the research of robot control system.As two basic schemes of improving robot intelligence,dynamic control can enhance the interaction ability between robot and environment,and visual servo can heighten the environment perception ability of the robot.In this paper,the visual servo control based on dynamic model of robot is studied,which aims to improve the visual servo performance of robot by considering the environment perception and interaction ability of robot.In the aspect of robot dynamic control,the first is to study the robot kinematic and dynamic modeling,and establish the robot dynamic model.In this paper,the robot ER3AC60 is taken as the research object.The kinematic model and differential kinematic model of the robot are established by DH method and vector product method,and the Jacobian matrix of the robot joints is obtained.The dynamic model of robot is established by Lagrange method and Newton Euler method.With known robot dynamic model but unknown parameters,the parameter identification method is further studied and the parameter identification experiment is carried out,in this way the estimated values of the unknown parameters in the dynamic model can be obtained.Firstly,this paper summarizes the commonly used dynamic parameter identification methods of robot,and selects genetic algorithm as the excitation trajectory optimization and parameter identification method.Then the excitation trajectory is designed and executed to collect the experimental data.Finally,according to the filtered experimental data,the unknown dynamic parameters are identified by genetic algorithm,and the accuracy of the identification results is verified afterwards.Combined with the dynamic model and parameter identification results of robot,the dynamic control method is studied and the dynamic control system is designed to realize the dynamic control of the robot.The most basic requirement of robot dynamic controller is to make the system level off to the expected value and keep stable;on this foundation,it is required to restrain the disturbance;furthermore,it is necessary to adjust the parameters adaptively to optimize the model.According to this progressive idea,this paper studies PID control,sliding mode control,sliding mode boundary layer control and sliding mode adaptive control respectively,and verifies the stability of the dynamic control system through simulation and physical experiments.In the aspect of robot visual servo,firstly,the model of robot visual servo system is established based on the modeling of visual servo system.Taking the general form of robot visual servo system as the research object,the camera imaging model is analyzed at the beginning to obtain the internal and external parameter matrix of the camera;then,according to the relative motion relationship between the servo target and the camera,the interaction matrix and the image Jacobian matrix are derived;after that,on the basis of camera model and image Jacobian matrix,the visual servo system is analyzed step by step from shallow to deep,and a visual servo system based on robot dynamic model is designed ultimately referring to the traditional model of robot visual servo.Ultimately,the visual servo system based on robot dynamic model is simulated and verified.The visual servo system is combined with the robot dynamic controller which is verified feasible to realize the visual servo control based on the robot dynamic model,and the results are compared with the visual servo based on kinematic to verify the feasibility of the algorithm.The experimental results show that the dynamic control system and the visual servo control system based on the dynamic model have good reliability in the actual robot,meet the requirements of robot control accuracy,and have academic value for further research.