Research On Visual Servoing Control For Constrained Manipulator

With the rapid development of of computer technology,automation technology and machine vision technology,robot technology,as a synthesis of the above technologies,has also achieved rapid development.The robot visual servoing control technology has been widely researched because of its ideal adaptability to the working environment.However,the manipulator visual servoing system is constrained in many ways.These constraints include physical constraints on the joint angle of the manipulator,the limited field of view of the camera,whether the image velocity can be obtained,the constraint surface that the manipulator works on,and the dead-zone nonlinear constraints.If these limitations are not fully considered,the control performance of the system will be seriously affected,or even the visual servoing will be failed.With the consideration of the above limitations,this paper studies the visual servoing tracking control for the constrained manipulator.The main research works are summarized as follows:(1)The fixed time control with limited joint angle tracking error for the manipulator is studied in view of the known working environment.A novel barrier Lyapunov function is designed to constrain the joint angle of the manipulator.Then,the large overshoot can be avoided and the joint angle of the manipulator will not exceed the actual physical limit.With the consideration of the uncertainties of the dynamics model and external disturbances,a fixed-time observer is designed to estimate the uncertainties and external disturbances.The non-singular fast integration terminal sliding mode method is used to realize that the tracking error of the system convergences to zero in a fixed time.By combining the fixed-time observer with the non-singular fast terminal sliding mode,the drawbacks of conventional control methods such as slow tracking speed and limited capability are solved.(2)For the unknown working environment,the visual servoing control for the manipulator with input dead-zone and limited field of view is studied.The adaptive method is used to compensate the effect of input dead-zone on the stability of the system.To sovle the problem of the limited field of view of the camera,the barrier Lyapunovfunction method is used to constrain the tracking error.Then,the image position is consrtained and that the feature points are always in the field of view of the camera can be ensured.Both steady symmetric and time-varying asymmetric constraints is considered,the controller is designed by using the back-stepping strategy,and the uniform ultimate bounded stability of the system is proved.(3)The visual servoing control problem with unknown image velocity information is studied.Under the depth-independent Jacobian matrix framework,a new image space observer is designed to solve the coupling relationship between dynamic model linearly parameterization and image velocity information.At the same time,the new image space observer does not use the inverse of depth estimation.Thus,the singular problem is avoided.An adaptive controller is designed to realize the visual control with unknown image speed information,and the asymptotic stability of the system is strictly proved.(4)The visual servoing and force control for manipulator with unknown constraint surface and input dead-zone is studied.The unknown constraint surface is linearly parameterized,and new adaptive laws are designed to estimate the unknown parameters online.A compensatory depth-independent Jacobian matrix is constructed to compensate the unknown time-varying depth information in the closed-loop system.At the same time,with the consideration of the input dead-time,a robust compensateory term is constructed to solve the influence of the input dead-time.By using Lyapunov stability theory,it is proved that both image position and force tracking errors can converge to zero asymptotically.