Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (9): 241-243.DOI: 10.3778/j.issn.1002-8331.2009.09.070

• 工程与应用 • Previous Articles     Next Articles

Multi-gases qualitative identification using SVM and data fusion

HUANG Wei-yong1,2,REN Zi-hui1,TONG Min-ming1   

  1. 1.School of Information and Electrical Engineering,China University of Mining and Technology,Xuzhou,Jiangsu 221008,China
    2.School of Information and Electronic Engineering,Xuzhou Institute of Technology,Xuzhou,Jiangsu 221008,China
  • Received:2008-01-28 Revised:2008-04-15 Online:2009-03-21 Published:2009-03-21
  • Contact: HUANG Wei-yong

多气体的SVM数据融合定性识别方法

黄为勇1,2,任子晖1,童敏明1   

  1. 1.中国矿业大学 信息与电气工程学院,江苏 徐州 221008
    2.徐州工程学院 信电工程学院,江苏 徐州 221008
  • 通讯作者: 黄为勇

Abstract: The traditional method for qualitative identification of multi-gases based on neural networks has the problems of over-fitting and poor generalization ability.In order to solve the drawbacks,this paper proposes a new method based on support vector machine(SVM) and multi-sensor data fusion,which uses multi-class classifier to fuse data of sensor array composed of several gas sensors,temperature sensor and humidity sensor,effectively eliminates the influence of ambient temperature and humidity on gas sensors,and reaches 100% qualitative identification rate.The experimental results show that the method is effective.

Key words: support vector machine(SVM), sensor array, data fusion, multi-gases qualitative identification

摘要: 针对基于神经网络的多气体定性识别方法中存在的过学习和泛化能力差的问题,提出了一种基于支持向量机(SVM)与多传感器数据融合的多气体定性识别方法。该方法采用结构化风险最小化准则的多类分类支持向量机对由多个气体传感器、温度和湿度传感器组成的传感阵列的数据进行融合,克服了传统方法的缺陷,消除了环境温度与湿度等因素的影响,实现了100%的定性识别率,实验结果证明了该方法的有效性。

关键词: 支持向量机(SVM), 传感器阵列, 数据融合, 多气体定性识别