Research On Real Time Detection And Correction Method Of Enameled Wire Winding Defects

With the continuous development of the economy and society,the investment in aerospace industry has gradually increased,which has created considerable benefits for China’s economic and social development.Solenoid valve is an important part of rocket engine.The winding quality of solenoid valve coil determines the stability and reliability of rocket power system.At present,the tension of the winding equipment of solenoid valve coil fluctuates greatly,there are many manual interferences and low information integration.The winding state of the coil can only be observed by manual monitoring.When there is a winding error,manual shutdown is needed to correct the current winding error,winding one layer and checking one layer.The work efficiency is low,the workload of the operator is heavy and the scrap rate of the coil is high.Based on the research of visual inspection technology,this paper designs a system which can detect and correct the winding defects of enameled wire in real time.The winding defects refer to the gap and lap error winding state in the winding process,and detect and correct the error winding state in real time,so as to improve the quality and production efficiency of the finished coil.The main contents of this paper are as followsFirstly,the winding process of solenoid valve coil is analyzed,the design scheme of visual inspection system is proposed,and the core hardware of the visual inspection system is selected;aiming at the distortion in the image,the camera is calibrated to ensure the best image quality;the actual size of a single pixel is calculated by linear regression.Secondly,according to the winding characteristics of enameled wire,a reasonable image processing flow is designed,and Otsu threshold segmentation,Gaussian filtering and other preprocessing algorithms are applied to the contour image of enameled wire;Canny edge detection algorithm and 1x3 template extraction contour point algorithm are studied respectively for the preprocessing effect image,and the experimental results of contour point extraction are compared.According to the characteristics of the coil contour curve,the Hough transform circle detection algorithm is applied to the contour curve.The experimental results are analyzed.The minimum points of the contour curve are segmented and the contour points fitting algorithm based on the least square method is used to realize the detection of the coil winding state.Combined with the practical application,according to the detection results of the visual inspection system,a reasonable correction method is proposed.The main control system controls the forward and reverse rotation of the winding motor and the forward or backward of the traverse motor according to the correction method to correct the current winding state.The correction method is tested.If there are still errors after more than 3 times of correction,the correction fails.The test results show that the correct rate of correction is 99.8%.The reason of correction failure is that the enameled wire will produce torsion during winding,which leads to correction failure.Finally,the stability test of the visual inspection system shows that the inspection system is stable and reliable.The test results show that the detection accuracy of enameled wire spacing is 99.6%.According to the orthogonal test of the detection system,the faster the winding speed is,the greater the influence on the detection accuracy is.