Research On Profile Error Evaluation And Uncertainty For Free-form Surface Based On Statistical Learning

With extensive application of parts with free-form surfaces in many fields such as automobile manufacturing,aerospace,shipbuilding,mold processing and many other industrial fields,the machining precision of this kind of parts becomes increasingly strict.Currently,how to realize the detection of free-form surfaces efficiently and accurately has played an important role in surface quality detection of products.In order to obtain the quality information of the free-form surface,the profile error evaluation algorithm its uncertainty estimation of free-form surface were studied in this paper.First of all,according to the classical definition of form geometric error,the mathematical model of free-form surface profile error evaluation was established in terms of the homogeneous coordinate transformation theory in rigid body kinematic and the minimum zone criterion.Least square method was used in this model to pretreat the measurement point set.Take the measurement points after pretreatment as the initial measurement data in fine positioning,thus a reasonable initial pose was provided for the fine positioning.Secondly,profile error evaluation of free-form surface using sequential quadratic programming(SQP)algorithm was studied.For surfaces whose only information given is its CAD model,the method of calculation the shortest distance from the measurement point to the free-form surface was proposed on the foundation of STL(STeror Lithography)model.For surfaces whose parameter expression is given,subdivision approach algorithm is adopted to calculate the point-to-surface distance.And the approximate linear differential movement model of signed distance is deduced which simplified the calculation procedure.Subsequently,optimal matching between measurement points and the design surface was implemented with sequential quadratic programming algorithm.In what follows,computer simulation was conducted by MATLAB and the contrasting results verified that the SQP-based profile error evaluation algorithm is feasible.Finally,the influencing factors of free-form surface profile error uncertainty were analyzed.Measurement uncertainty estimation model of free-form form surface was established in terms of the uncertainty transmission rules presented in Guide to the Expression of Uncertainty in Measurement.In the end,the above models and algorithms were verified through establishing the experiment system.