Research On Weld Recognition Of High-speed Rail Grinding Robot Based On Vision Sensor

The economic development has promoted the construction of high-speed railways,and the popularity of high-speed railways has promoted the development of cities around the railway.Seamless rail technology has made the speed of high-speed railways a new level.In modern societies,the method of artificially increasing the cost of labor and the relative backwardness of low automation equipment have left the method and technology for the grinding of rail welds unchanged for many years.With the increase in production requirements,industrial robots are playing an increasingly important role in all walks of life.In the repeated work environment of the welding track factory,industrial robots are more able to adapt to environmental requirements and work intensity requirements and complete the grinding work.The purpose of this project is to use machine vision technology to establish a visual system of rail welding robots,to achieve the positioning of the rail welds and the detection of the appearance characteristics,assist in the grinding of industrial robots for grinding,to achieve the automation of the grinding system,the goal of integration,and to obtain High precision,high accuracy grinding effect.The use of industrial robots to grind rails is a prerequisite for knowing the location and shape characteristics of the rail welds.The two rail connections are called welds.After the welded welds,there will be spilled welds between the two weld faces.In order to detect welds and polish them,weld identification is required.Weld identification includes the identification of weld locations and the detection of weld appearance characteristics,which are referred to as weld location and weld profile inspection.The purpose of the weld positioning is to find the relative distance between the robot end and the two-dimensional laser sensor,so that the robot end can drive the two-dimensional laser sensor to move over the weld seam.Only the two-dimensional laser sensor can accurately fall above the weld seam to accurately detect the shape of the weld seam and obtain the information of the weld seam profile in the weld seam.Weld seams were photographed using an industrial camera and the blue laser light source was used to obtain the corresponding picture.Digital image processing is performed on the image,and the improved canny edge detection technology is used to find the position of the weld in the image.The weld is calculated by converting from the position of the picture to a spatial position for grinding.The improved Canny edge detection technology uses the Otus operator to achieve automatic threshold acquisition in the detection process,replacing the traditional manual input threshold method.At the same time,morphological operations are used for filtering,so thatthe edges of detection results are smoother and more suitable for project requirements.The weld profile inspection includes the detection of a series of feature points and feature data,such as weld height detection at weld seams,weld spot width detection,and weld straightness detection.Firstly,a two-dimensional laser sensor is used to obtain the digital signal of the weld bead profile.Then the detailed information of the profile of the weld seam is obtained by wavelet decomposition and reconstruction.The detailed information is analyzed to obtain the edge points of the weld seam welding.The weld seam is calculated after calculation.Tumor profile data.For the detection of the effect of rail grinding,a two-dimensional laser sensor can also be used to obtain the surface data of the rails on both sides of the weld seam and fit them into a straight line,and calculate the straightness of the straight line between the two sides.After completing the above work,you need to transmit the various results obtained.This article uses Visual Studio to program and write interactive software in c++.The software has many functions such as calibration,connection,modification of parameters,positioning,shape detection,etc.The function is highly integrated and easy to observe and operate.The improved Canny edge detection operator of this project and the use of wavelet analysis method to analyze the characteristic points of the weld profile all play an important role in the process of weld sanding,not only improving the grinding speed,but also reducing the artificial labor of the grinding process.Realize the automation,integration,and modernization of the grinding system.