本文研究了多峰优化问题,利用梯度算子和筛选策略,得到了一种可求解多峰函数全部最优解的改进型遗传算法,数值模拟结果表明,该算法在处理复杂多峰函数优化问题时,局部搜索能力和克服过早收敛能力方面相对于传统遗传算法均有很大提高。
The article studies the multi-apices optimization problem. An improved genetic algorithm based on gradient operator and sifting strategy is proposed. The numerical experiments show that the algorithm exceeds standard genetic algorithm on the ability of local searching and premature convergence.