Research On Laser Remanufacturing Path Planning Algorithm Of Damaged Parts

The repair and remanufacturing of key parts of major equipment is one of the key research directions in the field of remanufacturing,and laser remanufacturing is increasingly used in the remanufacturing and repair of key parts of major equipment.However,today’s laser remanufacturing repair and repair efficiency is low,the repair process is not intelligent enough,and there are a lot of redundancies in the remanufacturing path.In order to improve the repair efficiency and quality of laser remanufacturing of damaged parts,and improve the intelligent level of repair,this topic focuses on the research of laser remanufacturing path planning algorithm,focusing on the identification,positioning and extraction methods of damaged parts,which is applied to the remanufacturing of damaged parts.The contents of the thesis as follows:Firstly,for the identification of part damage area,a multilayer perceptron damage recognition and positioning algorithm combined with curvature change factor is proposed,which obtains three-dimensional data of damaged parts through the threedimensional data measurement robot,preprocesses three-dimensional point cloud data,trains data sets,verifies data sets,trains models and verifies detection accuracy,and realizes the accurate identification,positioning and extraction of damage areas.The accuracy of this method in the damage detection of damaged parts reaches 99.60%,which provides a data basis for subsequent remanufacturing path planning.Secondly,according to the obtained three-dimensional point cloud data of the damage area,the point cloud slicing algorithm is used as a tool to obtain the initial path of the damage area milling process,and the RANSAC algorithm is applied to iteratively optimize the initial path.An optimal milling path optimization algorithm combined with milling cutter process parameters is proposed,combined with AABB bounding box theory,the initial milling path is optimized,and finally a complete milling processing path is generated,combined with the position information of the damage area,the program instructions that can be recognized by the milling robot are generated to complete the milling preprocessing operation.Thirdly,based on the damage area pretreated by milling,the laser repair path planning is carried out,and a laser repair path generation algorithm suitable for groove repair is proposed,and the theoretical lifting amount of multiple channels and multilayer Z-axis is determined by establishing a theoretical model,which is used as the theoretical basis for the lifting amount of laser head in actual laser repair.The collision detection in the laser repair process is studied,and the laser repair path is optimized and the laser repair robot movement is guided by simplifying the three-dimensional point cloud data,combining the point cloud collision theory,and comparing with other collision detection algorithms.Then,the laser repair area is post-grinded,and the grinding path generation algorithm is proposed,and the boundary grinding path and repair area grinding path are generated for the repair area and the repair boundary respectively,and the remanufacturing process of the damaged parts is completed by the grinding robot.Finally,taking the H13 mold steel damage substrate as the experimental object,the laser repair process parameters obtained by the pre-experiment are used to plan the whole process of laser remanufacturing of damaged parts,and the damage detection method and path planning algorithm proposed in this thesis are used to complete the laser remanufacturing of damaged parts,and the rationality and effectiveness of the remanufacturing strategy proposed in this thesis are verified by macro and micro tissue analysis.This thesis has a total of 81 charts,11 tables,and 90 references.