基于遗传算法的Rosenbrock函数优化问题的研究

OPTIMIZATION OF ROSENBROCK'S FUNCTION BASED ON GENETIC ALGORITHMS

本文利用遗传算法研究了极小化Rosenbrock函数的问题．较多的计算机模拟实验结果表明，遗传算法可以有效地解决这一问题．文中还分析了一些改进的遗传算法对于该问题搜索速度的影响，得到了适于解决此问题的合理的遗传操作，从而为有效地解决最速下降法所不能实现的某一类函数代化问题提供了一种新的途径．

This paper deals with the optimization of Rosenbrock's function based ongenetic algorithms. The simulated results show that the problem can be solved effectivelyusing genetic algorithms. The influence of some rnodified genetic algorithms on searchspeed is also examined. Some genetic operations suitable to the optimization technique areobtained, therefore, a novel way of solving a class of optimizations of functions that cannot be realized using the method of steepest descent is proposed.