基于频域分段-时域反演法的抽水蓄能机组大波动过渡过程水压脉动信号分析

抽水蓄能机组大波动过渡过程中水压脉动信号包含大量的毛刺,表现出瞬时大幅跳变的特点,严重影响水压脉动最值的分析选择。针对该问题,本文提出一种频域分段-时域反演法,首先将时域水压脉动信号进行傅里叶变换以获取水压脉动信号的频谱特征,在频域根据设置的截止频率对水压脉动信号滤波,经快速傅里叶逆变换反演出滤除指定频段后的水压脉动信号。通过在频域滑动截止频率,提取水压脉动极值-滑动截止频率关系曲线,分析不同频段对水压脉动最值的敏感性,揭示滤波频率对水压脉动信号最值的影响。该方法应用于工程实际抽水蓄能机组水压脉动信号分析,为水压脉动最值的确定提供了有力论证。

The Pressure Fluctuation Signal Analysis of Pump Hydraulic Turbine Based on Segment in Frequency Domain and Peak Analysis in Time Domain

During the load rejection process of pumped storage unit, there are many burr in the pressure fluctuation signal. The instantaneous sharp jump serously influences the analysis and selection of maximum or minimum pressure. Focus on the problem, the paper introduced an analysis method named Segment in Frequce Domain and Peak Analysis in Time Domain(SFDPATD). Based on this mehod, the pressure fluctuation signal was firstly carried out Fourier Transform and the spectrum was got. With setted cut-off frequency and inverse Fourier Transform used, the pressure fluction signal was filtered and the pressure fluctuation signal with setted frequency segment was got. With different cut-off frequency setted, the relationship between cut-off frequency and maximum or minimum pressure fluctuation was analized. Then, with the sensitivity analysis of different frequency bands related to the maximum or minimum pressure fluctuation carried out, effect of filtering frequency on pressure pulsation was displayed. With the mehod applied to pressure fluctuaion of pumped storage unit, it provides a powerful argument for the determination of the maximum value of water pressure pulsation.