LAUSR

Abstract For economic inspection of geometric errors on freeform surfaces, only a finite sample of points could be measured across the surface. Sampling strategy defines the number and distribution of this finite sample of measurement points. Only a well-selected sample of measurement points can accurately represent geometry of manufactured surface. Thus optimization methods could be employed to improve sampling strategies. In present work, a new hybrid sampling strategy is developed using Particle Swarm Optimization (PSO) method. The developed sampling strategy was compared with two existing strategies: (1) basic sampling strategy with uniform distribution in Cartesian space, and, (2) advanced adaptive sampling strategy developed by Yu et al. (2013). Accuracy and effectiveness of the developed method were verified through simulations. Case studies for several sample sizes showed that, the deviation of measured surface and its reconstructed model can be significantly reduced using the developed sampling method compared to the two aforementioned methods.