摘要
将混沌序列引入遗传算法中，获得了一种新的算法．该算法有效地克服了标准遗传算法中的“早熟”现象，并具有更快的收敛速度．计算机模拟验证了上述特征．同时还将改进了的算法用于多重小波神经网的训练．
In this paper, a chaotic sequence is inserted into the search procedure of genetic algorithm, which can overcome premature of the search by genetic algorithm and the speed of convergence is faster than standard genetic algorithm. Computer simulations show the effectiveness of this new algorithm. The new genetic algorithm is applied to train multiwavelet neural network for function approximation prediction.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
1999年第11期1-7,91,共8页
Systems Engineering-Theory & Practice
关键词
混沌序列
遗传算法
最优化
神经网络
chaotic sequence
genetic algorithm
optimization
multiwavelet
neural network
