| 高翔,陈向东,宋爱国,陆佶人.基于遗传算法的神经网络被动声呐目标分类研究[J].声学技术,1998,(4):169~172 |
| 基于遗传算法的神经网络被动声呐目标分类研究 |
| Study on neural network classifier of passive sonar target based genetic algorithm |
| 投稿时间：1998-06-22 修订日期：1998-09-02 |
| DOI： |
| 中文关键词: 声呐 目标分类 遗传算法 神经网络 |
| 英文关键词: sonar targets classification genetic algorithm neural network |
| 基金项目: |
|
| 摘要点击次数: 1320 |
| 全文下载次数: 1076 |
| 中文摘要: |
| 被动声呐目标识别系统中目标分类器的设计和训练是一项重要内容.本文设计了目标分类器的神经网络结构,提出了一种用改进的遗传算法训练神经网络分类器的新方法.最后,对海上实录的A、B、C三类目标噪声进行了分类识别,实验结果表明基于遗传算法的神经网络分类器比传统的基于BP算法的神经网络分类器泛化性能有明显提高. |
| 英文摘要: |
| The targets classifier is a key element in passive Sonar target recognition systems.In this paper,the structure of neural network targets classifier is designed.We proposed a novel method for training neural network targets classifier by using an improved Genetic Algorithm (GA).The targets classifier is used to classify three different classes of targets:A,B and C.The result of experiment shows that the preformance of GA based neural network targets classifier is better than that of Back propagation algorithm based neural network targets classifier. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
