Robotic grinding posture smoothing optimization for triangular mesh surface

At present, robotic grinding theory for mesh surfaces has been widely studied. However, existing methods often have the problem of poor smoothness, as well as neglecting the local differential characteristics of mesh surfaces. Therefore, in this paper, a robotic grinding posture smoothing optimization algorithm based on mesh surface and quaternion interpolation is proposed. Firstly, the local differential characteristics of mesh surface is reconstructed by estimating the normal vector of mesh surface and fitting the local geometric characteristics of mesh surface, in order to obtain the corresponding curvature information. Then, identification of key areas of the grinding process is conducted on basis of the obtained local differential characteristics, and key grinding postures are assigned at those grinding points with locally maximum curvature. Next, the technique of quaternion interpolation is performed among those assigned grinding postures to optimize motion smoothness and to improve the uniformity of grinding. The simulation experiment and practical grinding test have been carried out, and the results show that the proposed algorithm can generate smoother grinding postures along the path with less fluctuation, and can accommodate well with the changes of curvature on triangular surface, which ensures the smoothness of grinding and the uniformity of the surface after grinding.