Research On High-precision Reconstruction Of The Cross Section Data Based On Particle Swarm Optimization

"Digestion, absorption, innovation" is a proved effective way to the rapid development of new products. "Digestion and absorption," not only refers to the low level of similar shape, more importantly, a high-level parameters rehabilitation, in-depth understanding of the product content, and then the follow-up to further improve and innovate to provide the conditions. "Digestion and absorption," not only refers to the low level of similar shape, more importantly, a high-level parameters rehabilitation and in-depth understanding of the product content, and then provide the conditions to further improve and innovate. The reverse engineering modeling technology based on the features has been proved to be a kind of reverse modeling method, but in the reverse modeling process, reconstruction quality of section contour curve is still inadequate. Based on the maximum restore of the initial design intent and reconstruct of high quality section contour curve as the goal, a high-precision method based on Particle Swarm Optimization algorithm and curves meeting G~1 or G~2 continuity is proposed to solve the non-ideal cross section feature reconstruction result, which is due to poor extraction accuracy of segment points in reverse engineering, and the key is to extract precise segment points.In this paper, we introduced ways of section data acquisition and methods of data preprocessing. According to the discrete curvature of the cross section data information, extract segment points preliminary, and gives the feature recognition method based on section curve fitting. In this paper, in order to grasp the design features of the product, we used line, arc, and B-spline curve as the reconstructed sectional curve, in addition, B-spline curve we used in reconstruction can well capture the product characteristics, and the control point approach at least, to ensure a good smoothness of the curve. At the same time, we discussed the G~1, G~2 continuous and other constraints expressions between features.As to G~2 continuous cross-sectional reconstruction, because of its non-linear constraint conditions, we can solve nonlinear problems by linearization, but this method is relatively complex, usually prone to fitting failure. In this article, we used another way to gradually add constraint conditions. Based on preliminary reconstruction with G~1 continuous, we proposed an optimization model of adding G~2 continuous. First, reconstruction based on cross-sectional data with G~1 continuous, then inserted the optimal node selected by optimization models. Moved the fine control point according to G~2 continuous, which not only meted continuous condition, but also ensures the quality of cross-sectional data reconstruction. Then, based on the results of the above-mentioned reconstruction, used Particle Swarm Optimization algorithm with G~2 continuity constraints search precise segmentation, and improve the quality of cross-sectional reconstruction.The proposed high-percision reconstruction methods based on Particle Swarm Optimization algorithm of G~1 and G~2 continuous cross section data have been applied in the reverse engineering CAD software, STLViewer. Combined with the examples of section contour reconstruction of G~1 and G~2 continuous cross section data, then analyzed the quality of extracted segment point and the cross section contour curve reconstruction by using this method. Examples show that reconstruction by using this method well captured product’s original design intentions, improved the accuracy of segment point, and verified the feasibility and applicability of this method.