复杂背景噪声下风机叶片裂纹故障声学特征提取方法
发布时间：2019-03-23 17:25
【摘要】：针对大型风机叶片裂纹故障声学诊断问题,提出一种非接触式的叶片状态远程在线声学监测系统,给出了叶片裂纹故障的声学特征自适应提取方法.首先设计了面向复杂环境噪声的原始声信号预处理算法,然后采用1/6倍频程粗略刻画叶片声信号的频谱总体变化趋势,提取无量纲的倍频程能量比构造支持向量机分类器的输入特征向量,最后引入主成分分析法自适应的优化高维特征空间.风场实测数据验证了该算法的有效性.
[Abstract]:Aiming at the acoustic diagnosis of large fan blade crack fault, a non-contact on-line acoustic monitoring system for blade state is proposed, and the adaptive acoustic feature extraction method for blade crack fault is presented. In this paper, the preprocess algorithm of the original acoustic signal for complex ambient noise is designed, and then the frequency spectrum variation trend of the blade acoustic signal is roughly described by using 1 ~ 6 octave. The input feature vector of support vector machine classifier is constructed by extracting dimensionless octave energy ratio. At last, principal component analysis is introduced to optimize the high-dimensional feature space adaptively. The experimental data of wind field show the effectiveness of the algorithm.
【作者单位】： 北京邮电大学自动化学院;广东德风科技有限公司工程部;
【分类号】：TM315
，
本文编号：2446078
[Abstract]:Aiming at the acoustic diagnosis of large fan blade crack fault, a non-contact on-line acoustic monitoring system for blade state is proposed, and the adaptive acoustic feature extraction method for blade crack fault is presented. In this paper, the preprocess algorithm of the original acoustic signal for complex ambient noise is designed, and then the frequency spectrum variation trend of the blade acoustic signal is roughly described by using 1 ~ 6 octave. The input feature vector of support vector machine classifier is constructed by extracting dimensionless octave energy ratio. At last, principal component analysis is introduced to optimize the high-dimensional feature space adaptively. The experimental data of wind field show the effectiveness of the algorithm.
【作者单位】： 北京邮电大学自动化学院;广东德风科技有限公司工程部;
【分类号】：TM315
，
本文编号：2446078
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2446078.html
