支持向量机及其在密闭鼓风炉故障诊断中的应用.pdfVIP

2008 4 4 Journal of Chinese Computer ystems Vol. 29 No .4 2008 1,2 1 1 1 蒋少华 , 桂卫华, 阳春华, 谢永芳 1( , 410083) 2( , 512024) E-mail: sgjsh66@hotm : 针对密闭鼓风炉故障诊断中难以获得大量故障数据样本以及特征提取和诊断知识获取困难 不足, 提出了应用支持 向量机( VM) 进行故障诊断的新方法. 采用改进 “1 对其余”算法构建多 个VM, 利用可靠性数据分析技术中一些基本概念处 理原始样本数据作为特征向量, 输入到由多 个VM 构成的多类分类器中进行故障分类. 经实验证明, 该方法简单, 重复训练量 少, 训练、分类速度快, 准确度高. : ; ; ; VM : T P391:A : 1000-1220( 2008) 04-0777-05 Support Vector Machine and its Application to Fault Diagnosis for Lead-zinc Smelting Fur- nace 1, 2 1 1 JIANG hao-hua , GUI Wei-hua , YANG Chun-hua , XIE Yong-fang 1( School of Inf or mation S cience and Eng ineer ing, Central S outh Univ ersity, Chang sha 410083, China) 2( School of Inf or mation S cience and Eng ineer ing, S haog uan Univ ersity, S haog uan 512024, China) Abstract: Aiming at the difficulty in getting adequate fault samples, extracting eigenvectors and acquiring diagnosis know ledge - , in fault diagnosis for the lead zinc imperial sm elting furnace a novel method for the furnace fault diagnosis based on support vector machine ( VM ) is put forw ard. An improved one to ot hers algorithm is introduced to construct the multi-class VM . - classifier ome basic conceptions of dat a reliabilit y analysis are adapt ed to preprocess t he data as the input of the multi class classifier to identify faults. The method is simple and has lit tle repeated training amount . And the excellent performance on training speed and accuracy has been verified in the real application. Key words: support vector machine; dat a reliability analysis; fault diagnosis; mult i-class VM classifier 1