基于小波理论的汽轮发电机组碰磨故障诊断研究.pdfVIP

第十一届全国实验力学学术会议 大连 2005 基于小波理论的汽轮发电机组碰磨故障诊 断研究 高良丽 （大连理工大学，大连，116024） 摘要: 传统的傅立叶谱分析法适于稳态信号的分析，对非稳态信号，将其进行频率平均，不 能真实反映信号的特性。小波分析是一种时频分析方法，它在低频部分具有较高的频率分辨率 和较低的时间分辨率，在高频部分具有较低的频率分辨率和较高的时间分辨率，很适合探测信号 中的瞬态反常现象并展示其成分。本文运用小波理论对汽轮发电机组进行故障诊断。先在现 场用压电式加速度传感器对某发电厂正在运行的两台老汽轮发电机组进行振动测试，采集到 的振动信号通过模数转换(A/D)入计算机，运用小波分析进行特征提取并展示其细节成分。 结果表明该汽轮发电机组的汽轮机输入端轴系发生了严重的碰摩故障。 关键词: 小波分析；汽轮发电机组；故障诊断；碰摩; 傅立叶变换 Study on the Fault Diagnosis of the Serious Friction of Turbine-generator Sets Based on Wavelet Gao Liangli (Dalian University of Technology, Dalian, 116024) Abstract: Traditional fast Fourier transform (FFT) algorithm method adapts to analyzing steady state signals. As to the non-steady state signals, it averages their frequency, and can not show the real character of them. Wavelet is a method of time-frequency analysis. It possesses a higher frequency resolution and a lower time resolution in the low frequency part. In the high frequency part it possesses a lower frequency resolution and a higher time resolution. So it is well suited for probing the transient and abnormal phenomena, and shows their components. This paper diagnosed the turbine-generator’s faults by applying wavelet. First in a generating plant, the vibration signals of two old running turbine-generator sets are measured by applying piezoelectricity acceleration sensors in situ. The vibration signals are inputted into a computer after analog to digital (A/D). Applying wavelet the feature of vibration signals are extracted and the detail components are showed. The results show that there is serious friction in input-axis system of turbine. Key w