中文摘要：

针对自由曲面加工过程中的全局干涉问题,提出了一种基于自适应模拟退火遗传算法（SAGA）的快速全局干涉检验方法,将全局干涉检验转化为自由曲面上点到刀具轴线段的最短距离的寻优问题,获得了曲面上点到刀轴最短距离的直接求解方法,并基于矢量方法对干涉刀位进行了修正.设计了一种新的非线性自适应度函数,避免算法过早收敛于局部最优值.算法融合了自适应的交叉、变异算子以及模拟退火思想,经Schaffer函数测试其收敛概率约为99％.应用算例分析表明：自适应SAGA在全局干涉检验时计算量约为离散法的1/4,且避免了极值方法求解不收敛的问题,能直接计算出干涉点位置及干涉量,为全局干涉检验和刀位修正提供了科学依据.

英文摘要：

A method based on the adaptive simulated annealing genetic algorithm （SAGA） was proposed to solve the problem of global tool interference detection in free-form surface machining.The shortest distance between the tool axis segment and the point on free-form surface was solved by adaptive SAGA.Interferential tool position was adjusted based on the vector method.A new fitness function was designed for the algorithm integrated with adaptive crossover,mutation operators and simulated annealing （SA） idea.Its convergence probability tested by Schaffer function was about 99％.Application of adaptive SAGA showed that the amount of calculation was about 1/4 of the discrete method,thus the algorithm had better convergence properties.Using the algorithm,the global tool interference locations and values could be directly calculated.The algorithm provided a scientific basis for the global tool interference detection and tool position correction.