关键词: 自由曲面, 遗传算法, 适应度函数

Abstract: To select appropriate cutter size so as to realize free-form surface machining with no curvature interference, an Improved Genetic Algorithm (IGA) was proposed and applied in global maximum principal curvature calculating of free-form surface and maximum cutter size identifying with no curvature interference. A novel non-linear adaptive fitness function was designed according to principal curvature features, which could avoid trapping in local convergence. New selection rules of units to be copied were also given. Application of IGA showed that the global maximum principal curvature of free-form surface could be calculated effectively within 500 generations and the efficiency was higher than that of the discrete method. The cutter size selected according to the global maximum principal curvature was proper for free-form surface machining with no curvature interference.

Key words: free-form surface, genetic algorithms, fitness function, machining cutter