Research On Automatic Bending System Of Large Section Steel Based On Visual Measurement

In tunnels,bridges,housing construction,railways and other construction scenarios,the existing bending machine bends large section steel,there are local uncontrollable deformation of section steel,the swing range of the feed end section steel is too large and it is not easy to build a section steel production line,and the nonlinear rebound after the bending of the section steel needs to be corrected by repeated bending for many times to control the bending accuracy,etc.,this paper intends to use machine vision method to study and design a set of automatic section steel bending system that facilitates continuous feeding of section steel,realizes realtime detection and automatic adjustment of section steel bending radius.By understanding the principle and existing shortcomings of the existing section steel bending machine,the structural optimization design scheme of the section steel bending machine is completed,and the improved section steel bending machine defines the feeding route through multiple groups of clamping limit devices,so as to ensure that the section steel will not swing at the feeding end of the bending machine,which is convenient for continuous feeding;The bending of the section steel is carried out at the end of the bending machine,and because the section steel is only stressed on one side and the other side is free,the stress can be released in time,which largely avoids the local uncontrollable deformation,which can further reduce the lower limit of the bending radius of the section steel.By setting a machine vision-based section steel bending detection device at the discharge port of the section steel bending machine,the radius of the bent section steel is obtained in real time,and the feed amount of the jacking device is adjusted in time to complete the real-time automatic correction of the section steel curvature,so as to achieve the purpose of meeting the bending radius of the design requirements in one bend.In order to obtain high-quality images,the acquisition background,acquisition height,light source height and angle were analyzed and compared.In order to quickly extract the target information,the original image obtained was processed such as image smoothing,image binarization,morphological processing,edge detection,etc.,and the optimal algorithm was determined by analyzing and comparing the effect maps obtained by each processing algorithm.In terms of bending radius detection,firstly,according to the length characteristics of the contour,the target profile is separated by the bubbling sorting algorithm,and then the radius pixel value of the target profile is obtained based on the least squares method,and finally the conversion between the pixel radius of the section steel and the actual radius is completed by the proportional calibration method.A hardware experimental platform based on visual measurement of section steel bending system and a software development environment based on Visual Studio + OpenCV were built.The selection and installation of hardware such as light source,camera and lens of the visual inspection system was completed,and the software design and programming of the visual inspection system were realized.The bending test results show that the prototype built according to the design scheme of this paper can bend the section steel with a smaller bending radius.Visual inspection experiments were carried out on the standard test plate and section steel of different specifications,and the experimental results showed that the central angle corresponding to the arc in the collecting field of view needed to be greater than 30° and the visual inspection system had accuracy and stability.The stability of the vision inspection system was verified by multiple inspections of section steel of the same specification,and the test results were compared with the measurement results of the bow high chord length method.Therefore,the research in this paper has important theoretical support and engineering application value for the realization of large section steel automatic press brake.