Design Of Visual Positioning System For Evaporator Tube Hole

There are a large number of heat transfer tubes in nuclear power station,which are inevitably contacted with each other.In this way,heat exchange will inevitably damage the heat transfer tubes.In addition,considering that the liquid inside the heat transfer tubes of the evaporator is corrosive or even radioactive,all of the above factors will lead to the degradation of the heat transfer tubes,so it is necessary to carry out eddy current testing on the heat transfer tubes.Due to the large number of heat transfer tubes,it takes a long time to check the position of the probe manually in the process of inspection,and the efficiency is low.Therefore,it is necessary to develop a machine vision aided positioning system.Machine vision technology has a good applicability in various fields,at the same time,it has achieved rapid development in recent years.The pipe hole recognition and positioning system based on machine vision researched in this paper has greatly liberated human resources,improved the recognition efficiency,improved the accuracy and stability,and has extremely important theoretical and practical significance for the identification and positioning of evaporator tube holes.Firstly,the general scheme of the tube hole identification and positioning system for nuclear power plant is designed,then the overall framework of the tube hole identification and positioning system is given,and the hardware equipment such as light source,camera,lens and spider are analyzed and selected.The design of software structure and the description of development language and platform are given.Secondly,the calibration method of the binocular camera is studied,the camera imaging model is established,and the Zhang calibration method is designed to calibrate the camera.Accurate camera calibration is the basis of accurate positioning of the heat transfer tube hole.The related imaging model is established,and the mathematical transformation relationship of the three coordinate systems is studied.Combined with the actual needs of this project,Zhang’s calibration method is selected for camera calibration.Using Matlab toolbox to get the internal and external parameters of the camera,at the same time,the influence of radial and tangential distortion on the calibration is analyzed,and the radial distortion is calibrated.All of the above work lay the foundation for the positioning of the heat transfer tube hole.Thirdly,the image processing technology of heat transfer tube hole is studied.The mean square error MSE and PSNR of heat transfer tube hole pictures obtained by different filtering methods are calculated and compared,which proves the superiority of the improved median filtering algorithm.An improved Canny operator edge detection method based on adaptive median filter and maximum variance between classes is designed.The improved Canny operator edge detection method can better retain the edge information of the heat transfer tube hole image.The improved Canny edge detection method has good experimental results and meets the requirements of CGN.Finally,the identification and location of the heat transfer tube hole are studied.Hough hole detection is used to identify the tube hole.The coordinates of the center of the heat transfer tube hole and the radius of the tube hole are obtained by the least square method.Then,the single strain matrix is studied.However,because the pixels are not matched,the pixel coordinates can not be reconstructed.The sift and surf feature extraction algorithms are studied.The surf feature extraction algorithm is selected to extract the feature points of the heat transfer tube holes and complete the stereo matching.Combined with the principle of three-dimensional reconstruction,the actual coordinates of the heat transfer tube hole are obtained.Then,a number of experiments and error analysis are carried out to verify that the designed recognition and positioning system has good real-time performance and high accuracy.