Engine Oil Seal Assembly Dimension Detection System Based On Surface Structured Light 3D Imaging Technology

With the development of industrial manufacturing technologies to the direction of digital,networked and intelligent,the rapid,high quality and non-contact detection technology for the products has an important impact on their quality and production efficiency.The rubber oil seal is a precision sealing part on the engine crankshaft.If the assembly size deviation of this part is too large,it will lead to poor oil seal and oil leakage of the transmission parts,which will affect the service life of the engine.At present,the assembly size detection of the crankshaft oil seal mainly used the manual and contact measurement methods.The traditional manual size detection method has some problems,such as slow detection speed,estimation reading and missing detection.The contact measurement method may cause damages to the oil seal because it requires direct contact with the inner and outer rings of the oil seal.Moreover,the equipment needs to be cleaned and calibrated frequently,otherwise the measurement accuracy will be affected.In order to solve the uncertainty caused by manual detection and the problems existing in the measurement and detection during contact,and to meet the high efficiency and high precision detection requirements for the automobile engine assembly production line,this dissertation proposes an engine oil seal assembly detection method based on 3D surface structured light imaging technology.The method has the advantages of simple operation,fast detection speed and high precision.The main research content of this dissertation is as follows:(1)Via the review of relevant literature and the investigation in an enterprises,the detection requirements of the engine oil seal assembly detection system are defined,and the overall scheme of the assembly detection system is designed.Different oil seal detection schemes are drawn up according to the size and assembly characteristics of front and rear crankshaft oil seals,and the actual size and working space of the detection oil seals are combined.The selection of camera,lens,projector,robot arm and other hardware is finished.(2)The model of surface structured light 3D reconstruction system is analyzed,and the imaging principle of traditional model and improved model is introduced.A four-step phase shift extraction algorithm and a three-frequency heterodyne phase unwrapping algorithm are used to obtain the unwrapping phase,and the camera and projector imaging system are calibrated.The imaging accuracy of the three-dimensional monocular structured light imaging system was verified by the step block measurement.For the rear crankshaft oil seal detection,the coordinate transformation relationship between the robot and the imaging system is obtained by the hand-eye calibration method,and the coarse fusion of point clouds is realized.The improved ICP fine splicing algorithm based on the ISS-3DSC point cloud feature descriptor is used to realize the high precision splicing of the rear crankshaft oil seal.The multi-view splicing experiment is carried out on the micron-level matte ceramic ball to verify the splicing accuracy.(3)According to the characteristics of the oil seal point cloud data of the front and rear crankshaft,different filtering algorithms are used to preprocess it.The data of the inner and outer ring point cloud of the oil seal were extracted based on Euclidean clustering segmentation algorithm,and the plane fitting was carried out to calculate the height difference of the inner and outer ring of the oil seal.Finally,the front and rear crankshaft oil seal assembly simulation provided by the enterprise is used as the experimental object to verify the effectiveness of the algorithm and the stability of the detection system,which is in line with the production and detection requirements of the enterprise.