| Detection and assembly are two very important processes in the production process of the automobile glass lifter,especially with the continuously improving of automation of the automotive industry.Thus simply relying on artificial visual inspection and assembly is obviously unable to meet the development requirements of today’s automobile industry.In the assembly process of the automobile glass lifter,it is necessary to detect the position and size of the pulley,and assemble according to the detect results.Conventional detection methods such as photoelectric sensors can only detect the leak of the pulley and other parts,failing to detect details such as the size of the lifter components and the relative distance between components,which is unable to meet the requirements of fast and reliable manufacturing of the glass lifter.In particular,the detection and assembly of the pulley greatly affect the performance and longevity of the sheave lifter.Therefore,in this thesis a assembly line of the pulley for the sheave lifter based on machine vision is presented,which is designed to overcome the shortcomings of low efficiency,high cost and poor accuracy in the manual inspection and assembly method of pulley assembly,and realize the automatic detection and assembly of pulley assembly.In this thesis,the overall scheme of the embedded vision detection system is firstly designed,and the image processing algorithm of pulley detection is analyzed,especially the detection method of the pulley surface is analyzed,and the edge detection algorithm is analyzed emphatically.In order to improve the detection efficiency,the continuous detection algorithm is studied.Secondly,in the arrangement of assembly line,the mechanical structure of the assembly line of pulley assembly is designed,and simulated by ABAQUS software,which verified the influences of riveting speed and friction coefficient on the clamping force in the riveting process of pulley assembly and slideway.Finally,the control system of the whole equipment is designed by using PLC and ARM.Based on the Qt platform,the user interface is designed,which has the functions of controlling the visual system,displaying the detection data,real-time monitoring and mode switching,realizing the real-time monitoring of the production process.A long time continuous test is carried out in the factory’s actual assembly line.The test results show that the device developedin this thesis can reduce the wrong assembly rate of the pulley assembly,and can replace the existing manual inspection link and the assembly process,thus reducing the labor cost and improving the reliability of the product(false detection rate of 0.7 ‰).The detection system can meet the requirements of surface quality testing in other industries,and it has important theoretical significance and practical application value. |
