Research On Measurement Of Repeatability Positioning Precision Of Robotic Arm Based On Computer Vision

The requirements of modern industry for product manufacturing accuracy have become higher,and industrial robots play a very important role in the field of precision manufacturing.Traditional measurement methods have the disadvantages of low mechanical accuracy and incapable of real-time measurement due to mechanical contact.Location information.Therefore,many non-contact measurement methods have been developed one after another.Among them,machine vision measurement technology has become the focus of industrial robots for positioning and measurement research.This article mainly evaluates the repeatability accuracy performance of the XY-Theta platform.The working space of the platform is 300mm×300mm,which is controlled by a serial redundant robot arm with four rotating joints.The platform adopts two positioning strategies: coarse positioning mode and fine positioning mode.In the previous study,the repeated positioning accuracy of the XY-Theta platform was measured by two micrometers.This contact measurement method would cause interference to the test accuracy.In order to avoid the shortcomings of contact measurement method,this paper proposes a measurement method based on computer vision XY-Theta platform repeated positioning accuracy.Choose a high-resolution telecentric lens to avoid image distortion,and use the focusing technology of the vision system to calculate the best position for the camera to obtain a clear image.Collect multiple groups of images under different working conditions through the camera,and use the image registration method based on AKAZE features to perform non-linear local feature extraction on the images under each working condition and match withthe reference image.Corresponding to the relative displacement between the reference images,the relative displacement of each group of images is mapped to the Cartesian coordinate system to form a robotic arm positioning point cloud.Finally,the minimum encircling circle of the point cloud is calculated to obtain the repeated positioning accuracy under the current working conditions.The measurement accuracy of the proposed method is analyzed experimentally.The results show that the measurement error of the method is within 0.0176μm.The repeated positioning accuracy of the XY-Theta platform was measured using the proposed method.The results showed that the repeated positioning accuracy was3.2μm in the coarse positioning mode and 1.4μm in the fine positioning mode.The visual measurement measurement scheme is easy to operate,accurate and controllable,and can help to improve the performance of the robot,thereby improving the level of intelligent manufacturing technology.And can be extended to other 2-DOF robot platform accuracy measurement applications.