| 基于改进递阶遗传算法的RBF神经网络分类器 |
| |
| 引用本文： | 薛富强,葛临东,王彬.基于改进递阶遗传算法的RBF神经网络分类器[J].系统仿真学报,2010,22(2). |
| |
| 作者姓名： | 薛富强 葛临东 王彬 |
| |
| 作者单位： | 1. 解放军信息工程大学,郑州,450002;中国人民解放军94568部队,郑州,450047
2. 解放军信息工程大学,郑州,450002 |
| |
| 基金项目： | 河南省基础与前沿研究基金(082300413205) |
| |
| 摘 要： | 针对通信信号调制类型识别,应用递阶遗传算法动态确定径向基神经网络分类器结构。建立了新的适应度函数,该函数简单直观,待定参数少;同时结合相关联赛选择方法对选择算子进行了改进,增加了种群进化的多样性,避免了早熟收敛。仿真结果表明改进算法能更好地确定分类器结构,分类准确率更高。
|
| 关 键 词： | 递阶遗传算法 径向基神经网络 分类器 适应度函数 |
RBF Neural Network Classifier Based on Improved Hierarchy Genetic Algorithm |
| |
| XUE Fu-qiang,GE Lin-dong,WANG Bin.RBF Neural Network Classifier Based on Improved Hierarchy Genetic Algorithm[J].Journal of System Simulation,2010,22(2). |
| |
| Authors: | XUE Fu-qiang GE Lin-dong WANG Bin |
| |
| Institution: | XUE Fu-qiang1,2,GE Lin-dong1,WANG Bin1 (1.PLA Information Engineering University,Zhengzhou 450002,China,2.PLA 94568 Troop,Zhengzhou 450047,China) |
| |
| Abstract: | The hierarchy genetic algorithm was used to determine the RBF neural network structure in the modulation identification of communication signals.A new fitness function was given,which is simple and contains less undetermined parameters.Also,the correlation tournament selection method was used to improve the selection operator,which increases the diversity of population evolution and avoid the premature convergence.Simulation results show that the new algorithm can better determine the classifier structure a... |
| |
| Keywords: | hierarchy genetic algorithm RBF neural network classifier fitness function |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
