遗传多峰搜索

搜索多极值点问题是遗传算法研究领域内的一个新的方向。本文在传统遗传算法的基础上引入梯度算子、聚类算子和单亲繁殖（克隆，交将梯度平方和的倒数作为评价函数，用来搜索多极值点。实际测试结果表明：该算法对等高等距、不等高等距和不等高不等距情况都有很好的结果。梯度算子的引入，将梯度法和CA很好融合了起来，大大提高了GA的搜索精度和速度，收到了令工满意的效果。

Searching multi-maximum points using genetic algorithms

Searchingmulti-maximum is a new field in the research on genetic algorithms. B asing on traditionalgenetic algorithms, we improve it to search multi-maximum by introducing gradientcalculator and merging calculator and clone and using th e inverse of the square sum ofgradient as fitness function. The test results sh ow us that the present algorithm isefficient on all types function. Gradient ca lculator com bines GA and Gradient Algorithmsvery well and improves searching speed and prec ision obviously.