解调分析在机械设备故障诊断中应用的三个局限性研究  

Limitation studies on application of demodulation analysis in fault diagnosis

作　　者：丁康[1] 江利旗[2] 

机构地区：[1]汕头大学机械电子工程系,汕头515063 [2]重庆大学机械工程学院,重庆400044

出　　处：《汕头大学学报（自然科学版）》2000年第1期1-12,共12页Journal of Shantou University：Natural Science Edition

摘　　要：在具有齿轮、滚动轴承的机械设备故障诊断中，广泛使用解调方法分析诊断故障．从理论上分析．现有的解调分析方法存在的三个局限性：将不包括调制信息（故障信息）的两时域相加信号，也以其频率之差作为解调信号而解出；广义检波滤波解调分析中，由于取绝对值或检波过程可能产生混频效应；几种细化解调分析算法中，无法在细化分析的选抽时进行数字低通滤波，有可能会出现调制频率的高次谐波发生频率混叠而反折到低频部分的现象，提出了基于复解析带通滤波器的优化希尔伯特变换解调方法，将解调分析频率与滤波带宽建立严格的数学关系，可以克服三个局限性，将带通滤波、希尔伯特变换及选抽合为一体．运算速度快，是最佳的解调分析方法．Demodulation analysis is widely used in fault diagnosis of machines with gears and bearings. Thcoretically, the existing methods for demodulation analysis have three limitations. Firstly, two time- domain adding signals which have no fault information are demodulated as the frequency subtraction of the two. Secondly, during generallied detection-filtering demodulation analysis, aliasing effect occurs as a result of the improper selection of absolute value or in the process of detection. Thirdly, in several new Zooming demodulation methods, digital low-pass filtering cannot be carried out in the process of Zoom sub-sampling resulting in the reflection of higher harmonic components of modulating frequency to lower frequency ranges. In view of these limitations, this paper presents an optimized Hibert transform demodulation method based upon multiple analytic band-pass filter to overcome the three limitations under a strict mathematical relation between Zoom times and filter breadth. It is believed that the new method which combines band pass filtering with Hilbert transform and sub-sampling is an optimized demodulation method.

关 键 词：机械设备 局限性 机械振动 信号处理 解调分析 频谱分析 故障诊断 调制频率 

分 类 号：TH17[机械工程—机械制造及自动化] TP277[自动化与计算机技术—检测技术与自动化装置]