基于VMD-HHT方法的水电机组启动过渡过程振动信号分析研究

发布时间：2018-06-22 07:51

本文选题：水电机组 + 过渡过程　；参考：《工程科学与技术》2017年02期

【摘要】：随着中国水电机组数量及装机容量的不断增加,水电机组启动过渡过程故障问题也日趋明显,利用信号分析法对机组故障诊断得到广泛的应用。利用一种处理非线性非平稳信号自适应方法——变分模态分解(Variational Mode Decompostion,VMD),对葛洲坝某台机组启动过渡过程振动信号分析。首先,利用VMD方法对启动过渡过程仿真振动信号进行分析,验证VMD方法处理非线性非稳态信号的有效性。然后,利用该方法对机组启动过渡过程实测振动信号进行固有模态函数(IMF)的有效分离,通过Hilbert-Huang(HHT)变换对各IMF进行瞬时频率计算,得出相应的Hilbert谱,并将结果与经验模态分解(EMD)方法进行对比。从最终的振动信号分析中可以看出,通过VMD分解后的IMF可划分为3部分:第1部分主要为信号的发展趋势;第2部分主要为信号的基础分解,出现较强的规律性的"纺锤形"信号,且能够反映机组开机过程的振动变化规律;第3部分主要为干扰噪音或机组故障导致的高频特征信号,信号波形图无明显规律性,对应的Hilbert谱也明显具有主频波动小,幅值稳定等优势。结果表明:经VMD分解后,机组启动过渡过程振动信号各分量频率变化与机组转速时变规律吻合良好,能够有效提取特征频率,较EMD具有更好的自适应性,分析结果更加准确有效,能够更好地揭示水电机组过渡过程信号中的振动规律。
[Abstract]:With the increasing of the number and installed capacity of hydropower units in China, the problem of transient faults in the start-up process of hydropower units is becoming more and more obvious. Signal analysis method is widely used in fault diagnosis of hydropower units. The vibration signal of a unit in Gezhouba during start-up transition is analyzed by means of the variational Mode decomposition (VMD), an adaptive method to deal with nonlinear and non-stationary signals. Firstly, the VMD method is used to analyze the simulated vibration signals in the start-up transition process, and the effectiveness of the VMD method in dealing with nonlinear unsteady signals is verified. Then, the method is used to effectively separate the vibration signals measured during the start-up transition of the unit from the inherent mode function (IMF). The instantaneous frequency of each IMF is calculated by Hilbert-Huang (HHT) transform, and the corresponding Hilbert spectrum is obtained. The results are compared with the empirical mode decomposition (EMD) method. From the final vibration signal analysis, we can see that the IMF after VMD decomposition can be divided into three parts: the first part is the development trend of the signal, the second part is the basic decomposition of the signal. A strong regular "spindle" signal appears and can reflect the vibration variation law of the unit during start-up. Part 3 mainly deals with the high frequency characteristic signal caused by the interference noise or the fault of the unit, and the signal waveform diagram has no obvious regularity. The corresponding Hilbert spectrum also has the advantages of small main frequency fluctuation and stable amplitude. The results show that after VMD decomposition, the frequency variation of vibration signal in the start-up transition process is in good agreement with the time-varying rule of unit rotational speed, and the characteristic frequency can be extracted effectively, which is more adaptive than EMD. The analysis results are more accurate and effective, and can better reveal the vibration law in the transient process of hydropower units.
【作者单位】：四川大学水力学与山区河流开发保护国家重点实验室水利水电学院;西华大学流体及动力机械省部共建教育部重点实验室;
【基金】：国家自然科学基金资助项目(51309172) 西华大学教育部重点实验室开放基金资助项目(szjj2013-008) 重庆大学国家重点实验室开放基金资助项目(2007DA10512712405)
【分类号】：TV734

【相似文献】