基于点云的汽车喷涂路径规划

Vehicle Spraying Path Planning Based on Point Cloud

针对切片法生成的喷涂路径在边缘处喷涂不均匀,且对汽车复杂曲面规划时存在喷涂路径间隔不均匀等问题,提出一种基于复杂点云的改进投影法的路径规划方法。首先,利用改造后的FCM算法将点云分割成较简单的点云集;其次,通过平均法矢量和边界连接方式构建有向包围盒,基于包围盒生成点云切片,再对各切片采用加权投影法生成路径节点;最后,通过蚁群算法生成连续完整路径。仿真结果表明,在期望喷涂厚度为45μm的情况下对汽车前杠和后盖进行路径规划,前盖采用切片法的厚度误差为16.4μm,采用本算法的厚度误差为4.12μm,误差减小了74.8%,后杠采用切片法的厚度误差为27.12μm,采用本算法的厚度误差为4.86μm,误差减小了82%。实验结果证明该方法能有效规划出完整喷涂路径,在喷涂均匀性相对切片法有明显提升。

Aiming at the problems that the spray paths generated by slicing method are not evenly sprayed at the edges and the spray paths are not evenly spaced when planning the complex surface of automobile,this paper proposed a path planning method based on the improved projection method of complex point cloud.Firstly,the algorithm used the modified FCM algorithm to divide the point cloud into simple point clusters.Secondly,a directed bounding box was constructed by means of average normal vector and boundary connection.Point cloud slices were generated based on the bounding box,and then path nodes were generated by weighted projection method for each slice.Finally,the continuous complete path was generated by ant colony algorithm.The simulation results show that:When the expected spraying thickness is 45μm,the path planning of the front bar and the rear cover is carried out.The thickness error of the front cover using slicing method is 16.4μm,and the thickness error of the proposed algorithm is 4.12μm,which is reduced by 74.8%.The thickness error of the rear bar using slicing method is 27.12μm.The thickness error of the proposed algorithm is 4.86μm,and the error is reduced by 82%.The experimental results show that the method can effectively plan the complete spraying path,and the spraying uniformity is significantly improved compared with the slicing method.