Research On Casting Grinding Control Strategy Based On Image Positioning

Since the reform and opening up,China’s comprehensive national strength has increased significantly,its international status has been rising day by day,and its economic level has developed rapidly.China has become the world’s largest manufacturing country,but it is not yet a manufacturing power.With the implementation of "Made in China 2025" strategy,the combination of traditional manufacturing industry and electronic information technology will be the dominant direction of future science and technology.This paper selects complex castings in manufacturing industry as the research object,and develops an intelligent grinding system that can realize recognition,positioning and grinding.The system takes casting pictures by industrial camera,and transmits the casting pose and edge contour information to the upper computer through computer image processing.The upper computer converts the coordinate system and feeds back to the mechanical arm,and guides the mechanical arm to modify the grinding path adaptically according to the casting clip position to complete the burr grinding test of the casting.Due to the good lighting conditions in the test site,considering that each device in the system can not interfere with each other after installation,and can realize the communication protocol between the manipulator controller and the upper computer.In the experiment,the corresponding upper computer control program was written on THE PC side with the method of out of hand and Py Charm editor and Open CV visual library.Firstly,the original image of the casting to be polished was collected and preprocessed for analysis,and the ROI area was selected.The noise interference of the image was reduced by filtering,and the image of the target contour was enhanced.The identification and positioning algorithm based on template matching was used to fulfill the target requirements.Then,camera calibration and hand-eye calibration were carried out to determine the internal and external parameters of the camera and the spatial position relationship between the camera and the end actuator of the manipulator.The polishing trajectory was fitted by RANSAC algorithm and trial run was carried out with manual instruction to verify the accuracy of the calibration.Finally,the Socket communication was used to establish the communication mode between ABB industrial robot and the upper computer,and the position of each casting was photographed.The data was sent by the upper computer to the control cabinet through Socket communication to change the real-time grinding path and complete the casting grinding test.The robot arm intelligent system studied in this paper can meet the technical requirements in three aspects:grinding casting accuracy,image matching recognition rate and running time.