Design And Research On Vision System Of Automatic Spot Welding Machine For Circuit Breaker

As an important means of manufacturing and processing,welding technology has been widely used in industrial production.With the rise of intelligent manufacturing,the manufacturing industry has higher and higher requirements on the automation level of welding production.It is the trend to study the welding system with high precision,high efficiency and intelligent control.At present,for some non-standard,diversified parts of the welding,many enterprises or through manual or semi-automatic way.In order to solve these problems,such as low level of welding automation,low efficiency,unstable welding quality,high labor cost,poor working environment and easy to cause health problems.The main research content of this paper is a high flexible and automatic spot welding system based on machine vision control,which realizes the automatic ultrasonic spot welding and resistance spot welding of multi type parts of a circuit breaker,improves the welding quality and efficiency,and effectively saves the cost.First of all,aiming at the problem that the current spot welding machine can not meet the automatic welding of a total of 17 types of parts,this paper analyzes the characteristics of the parts and the related requirements of automatic control,and designs a visual system scheme suitable for this project.This paper introduces the selection method of main vision hardware such as camera,lens and light source,and completes the hardware construction of vision system.Based on Visual Studio 2015 platform,the UI of the control system is built by C #language.Then,aiming at the problem of low calibration accuracy of the edge position of the field of vision in the long distance and large field of vision in the hand eye calibration,the nine point calibration method is improved,which effectively improves the impact of lens distortion on the calibration accuracy of the field of vision edge;because the vision system has five camera stations,which involves more image processing work,in order to simplify the subsequent image recognition and detection algorithm,the method is improved Based on the analysis of image preprocessing algorithms such as de-noising,enhancement and morphology,the appropriate algorithm is selected to preprocess the image based on Halcon 12.0 image processing library,and the median filtering algorithm is improved to shorten the preprocessing time;in order to realize the automatic recognition and capture of multi types of bar,disc and welding piece parts,the image matching algorithm is analyzed and selected Appropriate matching algorithm can identify the type of the parts to be welded and determine the grasping position;through image matching to identify the bar annular groove,fit the edge of the annular groove,determine the axial position of the annular groove,eliminate the influence of fixture wear and other factors on the axial position of the annular groove,and realize the accurate welding positioning of the bar annular groove.Finally,the defect morphology of bar welding and disc welding is analyzed,and the detection of welding slot or disc solder joint after bar welding is realized by visual inspection.The welding yield and efficiency of the machine are verified in the actual welding production.Through the analysis of experimental data,the requirements of project automation and flexibility are met.