Research On Curved Surface Fitting And Error Analysis

In recent years, due to the rapid development of industrial technology, especially the progress of the digital control machine tool technology, from the traditional machine tool to today’s super machining center and modular machine tool, including CAD/CAM and other modern technology. The species of parts are becoming increasingly close to diversification, from simple geometric shape to complex surface parts, and the application of free complex surface is increasingly widespread. Meanwhile, the requirement of dimension precision is higher and higher; production and measurement of complex curved surface parts have become the most important part in the process of production and measurement, scholars faces with how to improve the problem of complex curved surface parts processing precision. In this paper, we mainly study the fitting and error analysis of complex surface. The main contents of this paper are as follows:In this paper, the surface fitting process is mainly divided into two steps, namely, first rough fitting and then fine fitting. Rough fitting makes use of the differential evolution method to find the approximate value of initial value; fine fitting uses coordinate alternation method to obtain the accurate value of initial value, the application of these methods mentioned above, has solved the shortcomings of the slowly iterative speed, and greatly promoted the progress of curved surface fitting. Later, the application of the principle of intersection has realized curve and surface curvature according to adaptive sampling.This paper introduces the concept of profile tolerance, the model of error analysis and evaluation. In the surface fitting process, in order to eliminate the influence of unfavorable factors, adopt the coordinate transformation, the alignment of local coordinate method, and apply the least square method to simplify the operational process, finally construct the theoretical basis for evaluation of the error by using the iterative closest point method and only applicable to the smooth surface of the projection method.In this paper, we develop an adaptive data sample collection; point planning and its related error analysis and evaluation, which is verified by several examples and the simulation analysis. By using the STL model to automatically pick up points, adaptive generation of data points and simulation of surface fitting based on these points, compared nearest point iterative method with projection method, and conduct error analysis, results demonstrate this method has a character of high efficiency and strong stability, and a good practical significance.