Research On Effects Of Freeform Surface Best Fit Algorithm On Uncertainty Of On-line Inspection

With the development of science and technology, Beautiful appearance and individuality of products are increasing demand. The application of freeform surface parts get more and more aboard in the aircraft, automotive, home appliances and other products. High-precision and high-efficiency inspection for the geometric are proposed. But the error inspection of its manufacture is very hard for its irregular shape. Firstly, using the on-line inspection system collects the measuring point set from the part surface and digitizes the part surface shape. Then using surface best fit algorithm match the measuring point set and nominal model. Finally evaluating the machining error of part based on the result of match.In the process of inspecting freeform surface by the on-line inspection system, the on-line inspection system is essentially a coordinate collecting device. The result of evaluation relies on proper number of measuring points and good sampling strategy. In this paper, the effect of surface best fit algorithm on the on-line inspection system is studied. The problem of matching measuring points and nominal model is resolved in this thesis. This paper establishes mathematical model to evaluate the uncertainty of surface best fit algorithm with the statistical bootstrap method. Using experimental method analyses the effect of number of measuring points and sampling strategy on uncertainty of on-line inspection system.Firstly, the best fit problem between measuring point set and nominal model is discussed. The mathematical model of surface best fit is established and the four critical issues in the process of surface best fit are described. The surface best fit algorithm based on iterative closest point algorithm is introduced. Using the given coordinate transformation matrix with computer simulation, the validation of proposed surface best fit algorithm is verified.Secondly, the uncertainty of surface best fit algorithm is evaluated. The conception of measurement uncertainty is introduced and the uncertainty of surface best fit algorithm is analyzed in this paper. The bootstrap method which is based on computational statistics is introduced. The surface best fit algorithm is combined with bootstrap method and the model for evaluating uncertainty of surface best fit algorithm is established. When the number of bootstrap replication is large, accuracy of the result is high, but the computational cost is very high. To balance the accuracy and computational cost, experiments were carried out. The experimental results show that when the bootstrap replication is taken 1000, not only the accuracy of result is high, but also the computational cost is low.Finally, using experimental method, the effect of surface best fit algorithm on on-line inspection system is researched. The freeform surface of an automobile hub die is inspected based on the on-line inspection system platform. Using the proposed method for uncertainty evaluation, the effect of surface best fit algorithm on the result of on-line inspection system is discussed.