一种新的求全局最优的遗传算法

利用近年来广泛应用的遗传算法，对积分-水平集求全局最优概念性算法的蒙特卡罗随机投点的实现途径进行了改进，从而得到一种新的求全局最优的遗传算法。文中算法设计了适合的初始种群的均匀设计取法、交叉算子、变异算子和选择算子，给出了测度的勒贝格计算方法，从理论上证明了算法的收敛性。最后，选用了标准测试函数并与其他算法进行了比较，从数值上验证了算法的可行性。

A new genetic algorithm for global optimization

The genetic algorithm being widely used in recent years is made use of, the Monte-Carlo implementation of the theoretical algorithm for the integral-level set global optimization is improved. Another new implementation approach that is the genetic algorithm implementation of the algorithm is proposed. The choice of the population based on the uniform design, the crossover operator, the mutation operator are designed and the algorithmic method of Lebegne are given. The global of this approach is proved. Finally, the standard test functions are selected and compared with other algorithms. The numerical results show that the approach is efficient.