Research On Robot Uncalibrated Visual Servoing Contril System

Robot technology is one of the most important invention in the 20th century and concentrated embodiment of automatization technology. Recently, Robot technology has become one of the representative strategic technologies in the high-tech field, which leads the fundamentally changes in production mode of traditional industry so as to have far-reaching influences on the development of humankind society. With the increasing needs for robot's dynamic characteristic, research on intelligent robot with sense become the one of the important content of high-tech plans for governments all over the world. In all kinds of sensors, vision sensor has become one of the most important robot sensors with its significant advantages of having large information quantity, broadly applicable fields and noncontact detecting. Robotic control system with vision sensor can enhance its adaptability to outer environments and extend its application area. It can be forecasted that the intelligent robot with sense will be applied more and more extensively. Traditional robot visual servoing control system is based on calibrated technologies, so that the control precision of the servo system depends largely on the precision of calibration. However in practice, a variety of reasons, limit the application of the visual servoing control method based on calibrated technologies to a great extent. Uncalibrated visual servoing has become a hotspot in the field of robot visual servoing control. Uncalibrated visual servoing means that vision control law is designed directly by the system state error from image plane without pre-calibrating the parameters of camera and robot, which controls the robot to make system error converge to a permissible region. The dissertation develop the studies of robot uncalibrated visual servoing control, which is still in primary original and exploring stage in the field of robot visual serving control and doesn't set up the uniform system info.The main work presented in the dissertation can be summarized as follows: 1. The development of robot visual servoing control are reviewed and the disadvantages of visual servoing based on calibration technologies are pointed out. The basic principal and development of uncalibrated visual servoing is described, and then the merits and faults of the exsiting uncalibrated visual servoing control method are analysed.2. The robot control method based on adaptive kalman filter is proposed. Robot control system is influenced inevitably by random noise during running. When statistical property of random noise is known, the conventional kalman filter can be adopted to hold back the effect of measurement noise on control performance. When statistical property of random noise is not available completely, the filter performance would drop down even to result in divergence. Based on the dynamics nonlinearities of robot manipulator, an adaptive kalman filter is designed and its application to robot control system is investiageted in simulation environment. The simulation results show that the designed filter can hold back the effect of measurement noise with unknown variance and mean on robot control system and the dynamic performance of control system is improved.3. An image-based robot visual servoing immune control method is proposed. Based on property of robot visual servoing control system,a visual controller is designed using pseudo-inverse of image Jacobian matrix and principle of immune feedback to be applied robot visual servoing system. 2D plane visual tracking experiments are carried out in a two link robot. Simulation results show that Pi-type immune controller could remove the error quickly for its immune feedback mechanism and control system performance is better than that of the conventional PID control.4. An experimental flat of robot uncalibrated visual servoing control system is set up. A MOTOMAN industry robot, a CCD camera and an image grabber card, together with PC host computer, construct a hardware platform of the theoretical research and simulation experiment for robot real-time uncalibrated visual servo control.5. The uncalibrated disturbance-rejection visual servo-control for a robot based on adaptive immune tuning method is proposed. The problem of uncalibrated disturbance-rejection visual servo-control for robotic manipulators is studied .A parameter tuning method for disturbance-rejection controller based on adaptive immune algorithm is proposed to overcome the difficulty in choosing parameters. The sufficient and necessary condition of stability for the nonlinear discrete second order extended state observer is proved, which is further used in the process of parameter tuning. The experimental results in a 6D0F industry robot verify the feasibility and validity of this method.6. A new robot uncalibrated visual servoing control method with double-loop structure based on auto disturbance rejection method is presented. When the object moves, the difference between estimated and real models would increase and estimation precision of extended state observer for estimating real-time disturbance would decrease. This would result in reduction of control performance, and so a new robot uncalibrated visual servoing control method with double-loop structure based on auto disturbance rejection is proposed. The online identification of the image Jacobian matrix using Kalman filter in the inner loop is performed for approximating the actual visual mapping model. The auto disturbance rejection controller is adopted in the outer loop, and a nonlinear extended state observer is used to estimate and compensate dynamically the system's disturbance for the estimated model. Tracking experiments with 2D moving object are carried out in a 6DOF industrial robot. Experimental results manifest the feasibility and validity of this method.7.A robot uncalibrated visual immune control based on least squares support vector regression (LS-SVR) method is put forward.Robot uncalibrated visual control based on intelligent algorithm is studied, and then, LS-SVR is applied to learn the complex nonlinear relationship between the robot's pose displacements and observed variations in the image features.The parameters of LS-SVR is determined by adaptive immune algorithm(AIA) plus 5-fold cross validation and visual controller is designed using principle of immune feedback. Visual positioning experiments with 2DOF～4DOF are carried out in a 6DOF industrial robot. Experimental results testify the feasibility and validity of this method.