大型复杂曲面零件加工余量均布优化问题研究

提出通过曲面的初始匹配和精确匹配来实现余量分布的优化,初始匹配决定后续算法的变量空间;精确匹配获得最佳的曲面匹配姿态及最佳的余量分布,精确匹配采用最小二乘方法构造评估函数,应用遗传算法和单纯形法混合寻优,直接对问题涉及的曲面匹配变换矩阵的6个未知量求解,应用结果表明该方法具有易地实现,算法稳定等特点,较好解决了大型复杂曲面类零件加工余量计算问题。

Optimization of allowance distribution on the workpieces with large sculptured surfaces in NC machining

To assure every measurement point to have even allowance, an optimization method is proposed for allowance distribution on large workpiece with complex surfaces, which consists of rough surface matching and accuracy surface matching. The rough matching decides the variable range of the follow-up algorithms, whereupon the accuracy matching acquires the optimal location and the optimal allowance distribution. To cope with the accuracy matching, a hybrid global optimization method is adopted. It shares tht advantages of both genetic algorithms (GAs) and simplex. The method recommended has been used in the 5-axis machining of both the large Kaplan and Francis hydro turbine blades. The practice shows that the method is easy to realize and the algorithm is steady and reliable.