多维函数优化的遗传算法研究

本文研究了求解多变量函数优化问题的遗传算法，在此算法中采用了十进制浮点数基因表示方法，并相应地提出了一种叠加零均值Gauss随机扰动的变异方法，研究表明，对于满足组件假说的多维函数优化问题，这种遗传算法具有较高的搜索效率．

A generic algorithm for optimization of multi-variable function

In this paper, an implementation of genetic algorithm for multi-variable function optimization is discussed. In the algorithm, the genes are coded into decimal floating number. Correspondingly, a mutation paradigm is proposed which creates an offspring vector by adding a Gausian random disturbance with zero mean to each component of the parent vector. Experimental studies show that this algorithm can perform efficient searching on condition than the dealt multi-dimensional optimization problem satisfies the building block hypothesis.