计算机与现代化

• 应用与开发 • 上一篇    下一篇

基于声学特征的化工防爆电机在线监测

  

  1. (东北石油大学计算机与信息技术学院,黑龙江 大庆 163318)
  • 收稿日期:2014-09-15 出版日期:2014-12-22 发布日期:2014-12-22
  • 作者简介:李军(1969-),男,吉林乾安人,东北石油大学计算机与信息技术学院副教授,硕士,研究方向:嵌入式系统; 李伟鑫(1989-),男,硕士研究生,研究方向:信息智能分析与处理; 李建平(1976-),男,副教授,研究方向:数据库及其应用。
  • 基金资助:
    黑龙江省教育厅科学技术研究项目(12511012)

Online Monitoring of Chemical Engineering Explosion-proof Motor Based on Acoustic Features

  1. (School of Computer & Information Technology, Northeast Petroleum University, Daqing 163318, China)
  • Received:2014-09-15 Online:2014-12-22 Published:2014-12-22

摘要: 防爆电机在正常工作时会出现各种各样的问题,为了尽量消除这些不确定因素,本文提出基于声学特征的故障监测技术。本文利用美尔频率倒谱系数进行特征提取,用以提高故障检测的识别率,并采用隐马尔科夫模型进行状态识别,使训练的故障模型与音频信号能够达到最佳匹配。

关键词: 防爆电机, 信号处理, 特征提取, 状态识别

Abstract: There are various problems appearing in normal work process of explosion-proof motor. In order to eliminate these uncertainties, this paper puts forward a fault monitoring technique based on acoustic characteristics. In this paper, we use the method of Mel frequency cepstrum coefficient for feature extraction, with the purpose of improving the recognition rate of fault detection. In addition, we adopt hidden Markov models to identify the state. Thus the training fault model and the audio signal will achieve the best match.

Key words: explosion-proof motor, signal processing, feature extraction, state recognition

中图分类号: