Research On Embedded Online Diameter Size Measurement System Of Axial Forgings

Axial forgings is the key components of machinery equipment, it is of huge socialvalue and economic benefits to comprehensively improve the production efficiency and rate of finished products of axial forgings. Axial forgings generally manufactured using the method of free forging, under the process, the diameter of which is an important basis to the suitable decision of reduction, forging force, forging times and so on. So it is of important significance for obtaining the final geometry size, saving energy and avoiding material-wasting to know the diameter size of axial forgings real time.However, contacting measurement by workers is widely used in free forging factories at the present stage.Take the foundation of existing research, this paper designed a set of embedded full automatic system aimed at solving measuring the diameter online to replace the workers measured method. The key points and innovative works of this dissertation are listed as follows:(1) Take shortages of existing forgings measuring technology into consideration, a system based on the comprehensive theory of laser ranging and computer vision was designed, which detected edges of forgings though laser sensor, measured the distance between two edges of forgings using displacement sensor,revised the initial diameter by photos containing edge information taken by the industrial camera.(2) The mechanical structure was designed and the model of primarily components was chosen. And this system is suit for measuring diameter under seven hundreds millimeters, and between the largest distance of twelve meters by calculation. If it measures at six to eight meters, and then the image identification of forgings edge is within two pixels, meanwhile take the inaccuracy of displacement sensor, the measuring accuracy of this system can be in one millimeter.(3) The control parts is based on embedded Linux operating system. In respect of hardware designing, this dissertation realizes the motor controlling through chip of CPLD which expanded by parallel bus of ARM. In the other respect of software, automaticmeasurement flow was programmed by Qt, the communication between ARM and CPLD was realized by Verilog HDL.(4) The self-adapting pictures processing program referring to kinds of axial forgings edges was designed by Open CV. In the stage of image processing, an algorithm of growing regional seed filling using self-adapting seed points was proposed. In the stage of image edge recognition, a mixed detection method synthesized Canny and Laplacian was innovated. In the computing of interval pixels within pictures, Hough transform added with threshold of levelness was adopted.(5) The result of experiment designed for this system shows that:measuring accuracy can meet the calculative anticipation, and this system is feasible for replacing manual measurement in forge plant.