电机故障诊断中噪声提取系统的设计  

Design of Noise Extraction System in Motor Fault Diagnosis

作　　者：肖峻[1] 黄海涛[1] 杨铮[1] 杜兆勇[1] 李光海[1] 

机构地区：[1]武汉理工大学机电工程学院,湖北武汉430070

出　　处：《压电与声光》2017年第1期131-135,共5页Piezoelectrics & Acoustooptics

基　　金：湖北省科学技术研究与开发基金资助项目(201397289)

摘　　要：设计了一种远场多电机噪声定向提取系统。该系统通过数字信号处理器DSP28335与六阵元麦克风阵列远场采集噪声信号,并结合多重信号分类(MUSIC)算法及简化的聚焦算法确定各电机的位置参数,最后通过阵列导向矢量的广义逆矩阵及位置参数,完成对各电机噪声信号的分离与提取。仿真实验表明,该系统简易可行,电机的位置误差在5°以内,提取的电机信号与原信号的吻合度较高。A directional noise extraction system for far-field multiple motor was designed in this paper.The system acquired the far-field motor noise through the DSP28335 and the six-array microphone.Then the location parameters of each motor was determined by combining the MUSIC algorithm and the simplified focusing algorithm.Finally,the noise signals of each motor can be separated and extracted based on the generalized inverse matrix of the array steering vector and the position parameters.The simulation results show that the system is simple and feasible,the motor localization error is less than 5°,and the extracted motor signal is highly identical with the original signal.

关 键 词：电机噪声 噪声提取 传声器阵列 MUSIC算法 聚焦算法 

分 类 号：TN911[电子电信—通信与信息系统]