| In the field of mechanical manufacturing, the constant enlargement of the market requirements and the rapid development of production technology lead to the increasment of demands for NC machining of complex shape part. The curvatures of free-form surfaces mostly present irregular distribution. As a result, it is difficult for feature information extraction and analysis. Moreover, it puts forward higher requirements for the related technology like NC machining planning, data management, and etc. Accordingly, this dissertation makes deep research on subdivision planning of complex surface, tool path generation, milling parameter intelligent strategy, and file management based on embedded system. Main research achievements are as follows:(1) An algorithm of boundary extraction and subdivision planning of complex parametric surface is proposed. On the basis of the complex surface feature recognition and feature surface patches division, further research of subdivision surface patch is carried out based on subtractive clustering and fuzzy c-means clustering algorithm. Firstly, the number and the center position of surface patches which can characterize a certain specific shape of surface domain are determined. Secondly, optimization treatment for the center position of clustering surface patches is carried out. Thirdly, surface boundary is constructed by the introduction of Voronoi diagram. Finally, an analysis example is given to illustrate the application of the method.(2) A strategy of tool path planning is proposed on the basis of parametric line method for surface patch machining. Binary tree subdivision algorithm is adopted. And then cutter path interval is calculated through the introduction of error compensation along feeding direction, and parameter increment value is obtained by introducing error compensation on the basis of studying the first order Taylor expansion. Based on this, the calculative precision of the interpolation point's parameter value of the complex parametric curved surface can be increased. Then by translation, rotation transformation for the interpolation point accurate, the accurate position incremental value is available. Finally, fairing operations are done on the straight convex polygon boundary.(3) A thorough research of intelligent decision and automatic planning of tool path of free-form surface NC machining is conducted. On the basis of studying the optimum strategy of interference-free tool combination for 5-axis NC milling of complex surface, an interference treatment strategy based on center points of surface patches is proposed. Furthermore, a structural model based on integrating artificial neural network and expert system is designed for establishing milling parameter intelligent decision system of free-form surface, which can realize intelligent decision of feeding and cutting parameters of free-form surface NC machining.(4) According to the requirements of massive data management and high reliability, technology of Flash file management system applied in embedded CNC system is researched. A strategy of wear-leveling for Flash sectors and damage management based on storage contents is proposed. With this strategy, rational utilization and effective management of the Flash storage space can be achieved. In addition, a data management strategy based on doubly-linked list structure for file information is put forward, on the basis of which, program data are effectively maintained. Finally, aiming at program storage for large-scale parts and distributed processing, serial communication between CNC system and PC is realized. A CNC system prototype of TDNC-L4 is developed based on the foregoing research achievements.
|
PDF Full Text Request