基于奇异值分解的水流量标准装置状态监测方法研究

Research on State Monitoring Method of Water Flow Standard Facility Based on Singular Value Decomposition

提出了基于奇异值分解的流量计输出信号分离方法,对流量信号、流量波动信号和噪声进行分离,准确获得装置管路中不同频率的流量波动信号及流量波动相对幅度,实现装置状态监测。通过仿真实验准确分离出了不同频率的流量波动信号,而且即使噪声方差为1时的最大幅值误差也只有0.0681 m^3/h。最后通过实验分离得到:幅度分别为0.73%和0.25%的流量波动信号与管路中的水泵失稳和阀门摆动引起的流量波动频率一致。仿真与实验结果验证了所提方法的可行性。

A signal separation method of flowmeter based on singular value decomposition is proposed to separate the flow signal,flow fluctuation signal and noise,and accurately obtain the flow fluctuation signals with different frequency and the relative amplitude of flow fluctuation in the pipeline of the flow facility,so as to realize the condition monitoring.The simulation results show that the flow fluctuation signals with different frequencies are separated accurately,and the maximum amplitude error is only 0.0681 m^3/h even when the noise variance is 1.Finally,the flow fluctuation signals with relative amplitude of 0.73%and 0.25%are separated by experiment,whose frequency are consistent with the frequency of flow fluctuation caused by pump instability and valve swing.Simulation and experiment results verify the feasibility of the proposed method.