基于小波分析及振动信号灰度矩的水电机组振动区建立方法研究

通过对仿真振动信号进行连续小波变换,选择合适的小波分解尺度和小波类型,得到能够反映振动特征的小波系数图。由于系数图不能定量地反映机组振动强弱,需要人工介入辨识;于是在信号的特征提取中,引入"灰度矩",并把一阶矩作为定量指标,发现灰度矩数值能定量表征振动信号强弱。基于此理论方法,结合云南某水电厂的实际振动数据,提出一种建立水电机组振动区的新方法;并以上导水平摆度的振动数据为例确定出该机组的振动区;该方法比传统的仅仅通过振动峰峰值确定振动区方法更加准确,没有遗失振动的大量信息;而且以机组自身振动为参考标准,更有说服力。

A research method of hydropower generating unit vibration area establishment based on wavelet analysis and vibration signals gray moment

The simulation vibration signal continuous wavelet transform to select the appropriate scale wavelet decomposition and wavelet types, the resulting vibration can reflect the characteristics of wavelet coefficient map, since the coefficient map can not be quantified reaction unit vibration strength and requires manual intervention to identify. So in the signal feature extraction, the introduction of the "gray moment" and the first moment as a quantitative indicator, found gray moment value can quantitatively characterize the vibration signal strength. Based on this theory, the actual vibration data of a hydropower plant in Yunnan, the proposed new method of establishing hydroelectric generating vibration area and vibration data above guide level swing as an example to determine the vibration area of the unit. This method than the conventional kind of vibration just by determining peak to peak vibration zone method is more accurate and does not lose a lot of information vibration and vibration to the unit itself as the reference standard, more convincing.