多亲杂交算子及其在连续优化中的应用

提出了一种新的杂交算子，即基于邻域的多亲杂交算子，这种算子不仅呈现出多亲杂交的形式，而且内在隐含变异算子的特征。另一方面，为了加快收敛速度，引入了一种自适应的机制根据演化的进度调整邻域范围。实验表明，使用该算子的实数编码遗传算法在优化高维连续函数时是可行而有效的，尤其在BUMP问题的求解上有较大突破。

Multi-parent Crossover and Its Application in Continuous Optimization

This paper presents a novel crossover operator, neighborhood-based multi-parent crossover operator (NMPC), which not only takes on the form of a crossover operator, but also has the characteristics of a mutation operator. On the other hand, to enhance the convergent rate, it introduces a self-adaptive mechanism of adapting the range of the neighborhoods according to the evolutionary progress. Through some experiments, the proposed algorithm using the new operator is both feasible and efficient and especially a breakthrough is made in solving the BUMP problem.