The trajectory planning system for spraying robot based on k-means clustering and NURBS curve optimization

Abstract

In the spraying industry, the trajectory planning is the core part of industrial robots because model variety and shape of work-piece surfaces become the core problem of trajectory generation. In order to quickly plan the spraying trajectory and process of industrial robots, this paper proposes a trajectory generation model based on k-means mean clustering and a fitting optimization method based on NURBS curve. The model obtains point cloud data, classifies and iterates the curvature according to the curvature threshold, and then plans the trajectory of each region according to the height and Angle of the spray gun in the spraying experience base to obtain the regional spraying trajectory points. The NURBS curve is used to optimize the trajectory. In addition, on the basis of the above, a method based on infrared imaging target detection technology is proposed to quickly identify the acquired trajectory information and convert it to the coordinate system suitable for spraying robot operation.