• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2010, Vol. 32 ›› Issue (9): 53-56.doi: 10.3969/j.issn.1007130X.2010.

• 论文 • 上一篇    下一篇

可增量学习的水下航行器噪声源识别中聚类算法研究

高志华1,贲可荣1,章林柯2   

  1. (1.海军工程大学计算机工程系,湖北 武汉 430033;2.海军工程大学振动与噪声研究所,湖北 武汉 430033)
  • 收稿日期:2010-03-16 修回日期:2010-06-18 出版日期:2010-09-02 发布日期:2010-09-02
  • 作者简介:高志华(1979),女,湖北武汉人,博士,讲师,CCF会员(E202011937M),研究方向为人工智能与机器学习。
  • 基金资助:

    国家自然科学基金资助项目(50775218)

Research on the Clustering Algorithm of the ClassIncremental Learning Model for Underwater Vehicle Noise Source Recognition

GAO Zhihua1,BEN Kerong1,ZHANG Linke2   

  1. (1.Department of Computer Engineering,Naval University of Engineering,Wuhan 430033;
    2.Institute of Noise and Vibration,Naval University of Engineering,Wuhan 430033,China)
  • Received:2010-03-16 Revised:2010-06-18 Online:2010-09-02 Published:2010-09-02

摘要:

水下航行器的噪声源识别具有训练样本有限,存在偶发或突变噪声源等特点。本文针对这些特点,在具有增量学习能力的水下航行器的噪声源识别系统架构下,提出了一种参数自适应可调的基于密度的聚类算法。实验表明,该算法可以有效避免基于密度的聚类算法的参数敏感性对聚类结果的不良影响,在无监督情况下对水下航行器的机械噪声源样本进行有效聚类。通过该聚类算法标注后的样本可直接作为具有增量学习结构的分类器的训练样本,节省了时间和系统开销。

关键词: 噪声源识别, 增量学习, 聚类算法

Abstract:

The  underwater vehicle machinery noise source recognition features that  the training samples is limited and have abrupt noise samples. Based on these characteristics,this paper proposes a densitybased algorithm which is parameter adjustable. And this novel algorithm is an  important component of the underwater vehicle machinery noise source recognition system with incremental learning ability. The experimental results show the new algorithm can avoid the parameter sensitivity of DBSCAN. Labeled samples by this algorithm can directly be used as the classifier training samples,saving lots of time and system resources.

Key words: noise source recognition;incremental learning;clustering algorithms