摘要
电机出厂检测非常重要。为了确保质量控制,目前电机厂多是通过操作者听电机声音判定噪声故障。而该文将研究模式识别技术在小型电动机生产线上电机故障检测中的应用。由于工业现场环境,系统首先采用小波分析对振声信号进行消噪,提取有用信号。再利用小波技术多分辨率特点和小波能谱熵提取故障信号的特征信息,最后结合概率论参数区间估计法获得小波熵带,对故障电机自动识别。
It is very important to inspect finished motors.In most cases,an operator listens to the motor and audibly detects the noise faults to guarantee on-line quality control.This paper dealed with application of pattern recognition techniques to perform an automatic identification of noise faults in small motors applied to production line.Due to noisy conditions in industrial environments,this paper described a preprocessor based on wavelet analysis to suppress undesired noise.Moreover,faults feature is extracted by multi-resolution characteristics of wavelet technique and wavelet entropy spectrum theory.At last,wavelet entropy inter-zone established based on inter-zone estimate is used to identify defect motor automatically.
出处
《微电机》
北大核心
2011年第7期101-103,共3页
Micromotors
关键词
电机故障
小波
模式识别
motor fault
wavelet
pattern recognition
