Optimization of point cloud registration algorithm in robotic belt grinding of turbine blade

CHEN Wei，YAN Si-jie，ZHANG Jia-jun，ZHANG Hai-yang

（School of Mechanical Science & Engineering，Huazhong University of Science &Technology，

Wuhan 430074，China）

Abstract：Aiming at solving the problem of calculating the workpiece coordinate system in the robot grinding paths and calibrate the clamping error，the 3D point cloud registration technology was applied to the robotic belt grinding system for turbine blades. After acquiring the point cloud of workpiece by 3D laser scanner，the combination of principal component analysis（PCA）algorithm and improved iterative closest point（ICP）algorithm were employed to match the scanning points and discrete points of part model，then the workpiece coordinate system was obtained and the clamping error was known. Simulation and experimental results indicate that the registration speed and accuracy of the optimized algorithm are fully improved，and the machining accuracy and surface quality can meet the actual machining requirements.