Compressor Valve Assembly Robot Visual Guidance Positioning Research

With the development and maturity of machine vision technology and image processing technology,as well as the rise of labor costs,image processing technology and machine vision technology are increasingly applied in the field of intelligent assembly of industrial robots.Aiming at the automatic assembly process requirements of typical air valve work piece in compressor,this paper systematically studies the positioning and assembly technology based on visual guidance of industrial robot to realize the precise positioning of air valve parts of compressor.On the basis of detailed analysis of the visual positioning and assembly requirements of the assembly robot,this paper designed the overall visual positioning and assembly scheme in line with the laboratory environment.The "hand-eye" vision guidance system based on monocular vision was determined.According to the precision requirements of the vision system,the kinematics of the robot,the imaging model of the camera,the internal and external parameter model and the hand-eye calibration principle were discussed in detail.The camera calibration parameters and the hand-eye calibration conversion matrix were obtained.According to the geometric features of the valve to be identified,the image processing operation of the compressor valve is completed by combining the principles of image processing,threshold segmentation,morphological operation,edge detection and other image processing algorithms.For the realization of fast and accurate identification,positioning and assembly of work piece.Based on improved Hough transform and least square ellipse fitting,an improved identification and location algorithm for ellipse center with strong robustness and short response time was developed.Based on the independent research and development of the basic image library of OpenCV,Visual c++ is applied to realize the identification and positioning of the air valve contour,and the positioning result of the compressor air valve in the image coordinate system is converted to the description of the robot base coordinate system through coordinate transformation.Finally,ER30 industrial robot isused as the platform to build a visual guidance system test platform for assembly robot based on the above scheme.Using the industrial robot vision system,the positioning accuracy and efficiency of the compressor valve work piece were tested,and the visual assembly of the valve was tested.The accuracy and effectiveness of the visual guidance positioning system are verified.This subject theory and practice combine,belong to actual project.This study overcomes the difficulties in manual assembly of valve components of 2D80 reciprocating compressor,such as narrow operating space,difficult operation and high labor intensity.Stable assembly precision and quality,reduce repair rate and scrap rate,improve the competitiveness of enterprises.