摘要
时域平均法是故障诊断中常用的非常有效的信号处理方法,主要特点是可以从混有噪声干扰的信号中提取感兴趣的周期性信号,同时抑制不相关的其它周期成份。然而时域平均存在两方面明显的不足:一方面,键相信号的拾取常受到应用环境和条件的制约难以实施,即便获得键相信号,也往往由于周期截断误差的存在使时域平均方法不能取得满意的效果;另一方面,时域平均要求大量数据的存储、运算,这给时域平均的应用也带来一定的不便。为了解决上述问题,本文根据时域平均的幅频特性提出采用FIR多带滤波器实现时域平均功能的方法,该法有效避免了传统方法需采集键相信号的不足,很好的扩大了时域平均在实际工程中的应用。最后,将传统的时域平均方法和利用多带滤波器实现时域平均功能的方法进行了分析比较,验证了本文方法的有效性。
The TDA is a very useful signal processing method commonly used in mechanical fault diagnosis. It can be used to extract the specific periodic components from noisy signals and at the same time to suppress other periodic components. But the TDA has two questions,on the one hand,the acquisition of key-phase signal is difficult and the treatment processing is usually influenced by both environment and condition factors. On the other hand,it is also difficult to obtain the satisfied results because of the existense of periodic truncation error. In order to solve these questions,a multi-band FIR filtering based TDA method was proposed. In the method,the key-phase signal is not necessary,which avoids effectively the shortage of traditional methods. This makes it suitable to a wider class of engineering applications.The performances of the traditional TDA method and the multi-band FIR filter method were compared and analyzed. The results show the effectiveness of the new method.
出处
《振动与冲击》
EI
CSCD
北大核心
2017年第21期192-195,247,共5页
Journal of Vibration and Shock
基金
国家自然科学基金面上项目(51379189)
中央财政支持地方高校发展专项海洋工程装备创新团队
浙江省科技厅重大专项社会发展项目(2013C03031)
关键词
时域平均方法
多功能滤波
故障诊断系统
time-domain average method(TDA)
multi-functional filter
fault diagnosis system
