Research On Accurate Motion Control Of Manipulator Based On Visual Servo

At present,in view of the fact that the application of robot in China is still at the stage of teaching-playback,robot is required to pre-determine its trajectories and to have a high positional accuracy about surrounding environment,which greatly reduces the flexibility of the production line.Under this background,the visual sensor,because it provides a wealth of target information as feedback,is gradually applied to industrial production.And among those applications,robot motion control,based on the visual servo,becomes one of the most extensive applied research themes.Taking the UR3 manipulator as the research carrier,this paper will discuss the precise motion control research of the manipulator based on the visual servoing by the aid of industrial camera.The main contents of this paper are as follows:(1)On account of the objective of this paper,a binocular-mixed visual servo control system is proposed.In view of the advantages of Eye-to-hand/Eye-in-hand installation methods,Eye-to-hand installation method is applied to provide the target object space three-dimensional coordinates and to guide the manipulator to approach the target object based on the location of visual control while Eye-in-hand installation of the camera is applied to accurately locate the target and to accurately capture,place the target based on mixed-visual servo control.The experimental results show that the visual servo system has high stability and practical effect when it works.(2)To overcome the insufficient that the traditional second-order central moments can't calculate the image orientation angle accurately,a new algorithm based on image is proposed to directly measure the target in image processing and to accurately estimate the direction angle of the cube from the segmented image.(3)Since visual servo is a multi-input and multi-output nonlinear system,a fuzzy visual servo control algorithm is designed based on parameter adaptive in the research,namely,with the feedback image and the expected image characteristic error as the input,the robot pose increment signal is obtained by fuzzy reasoning and the joint angle of the manipulator is calculated by inverse kinematics.Comparing with the conventional fuzzy visual servo control algorithm,the fuzzy visual servo control algorithm based on parameter adaptive,according to research simulation results,has better control effect according to research simulation results.(4)In order to verify the effect of this algorithm,the control software is written on the VS2010 platform using C++ language.The experimental results show that success ratio for placing of the proposed algorithm based on parameter adaptive fuzzy visual servo control algorithm is 6.7% higher than that of the conventional fuzzy visual servo control algorithm and the time it takes to grasp and place two cubic blocks has saved 13 s in average,which verifies the superiority of the control algorithm proposed in this paper.In the experiment,however,the system has a certain degree of hysteresis which results from the time-consuming image processing.Thus,it is necessary to study more efficient and intelligent image processing and image classification algorithms later in order to improve the motion control performance of the system.