Method for automatic glue-spraying of stringer and inspection of glue-spraying quality

A method for automatic glue-spraying of a stringer and inspection of glue-spraying quality, comprising:

• (1) collecting data of a 3D point cloud of an assembly of a stringer and a skin and pre- processing the collected data of the 3D point cloud prior to glue-spraying;
• (2) extracting feature points of an intersection line of a site to be glued of the stringer and the skin;
• building a minimum spanning tree;
• and connecting the feature points to obtain an initial feature intersection line between the stringer and the skin;
• wherein step (2) comprises:
• (2.1) extracting the feature points of the intersection line of the site to be glued of the stringer and the skin based on extraction of feature points from a point cloud data model;
• wherein in step (2.1), the feature points of the intersection line of the site to be glued of the stringer and the skin are extracted by a K-means clustering method based on Gaussian mapping through steps of:
• randomly selecting a 3D point in the 3D point cloud as a target point
• performing a K-nearest neighbor search on the target point
• subjecting a unit normal vector of a triangle set composed of the target point and its neighbor points to Gaussian mapping;
• selecting silhouette coefficient as a clustering validity index to determine an optimal number of clusters;
• and
• obtaining the feature points in the 3D point cloud model as the feature points of the intersection line of the site to be glued of the stringer and the skin according to clustering distribution of different patches;
• and
• (2.2) building the minimum spanning tree based on the extracted feature points;
• and connecting the feature points according to the minimum spanning tree;
• (3) optimizing the initial feature intersection line, and transforming the optimized feature intersection line to a coordinate system of an end effector of a glue-spraying robot to obtain an actual glue-spraying trajectory;
• wherein step (3) comprises:
• (3.1) optimizing the initial feature intersection line between the stringer and the skin to remove redundant small branch structures;
• and
• (3.2) transforming the optimized feature intersection line to the coordinate system of the end effector of the glue-spraying robot to obtain the actual glue-spraying trajectory according to a determined calibration relationship;
• (4) after the glue-spraying, collecting point cloud data of a glue sprayed on the stringer and the skin;
• calculating size information of the glue and determining a position of defects on the glue based on the point cloud data of the glue;
• and performing secondary glue-spraying on the defects of the glue.