微弯型科氏质量流量计测量气-液两相流研究

Study on measuring gas-liquid two-phase flow with micro-bend type Coriolis mass flowmeter

与普通U形和△形科氏质量流量计相比，微弯型科氏质量流量计固有频率更高、相位差更小，测量气-液两相流时误差更大。为此，设计气-液两相流实验方案，采用课题组研制的科氏质量流量变送器进行气-液两相流实验，采用BP人工神经网络对测量误差进行建模，得到误差模型，实现对气-液两相流测量误差的在线实时修正。实验结果表明，当密度降在0％～30％范围内变化时，通过在线修正，气-液两相流测量误差从原来的最大为-50％减小到-5％～3％以内，取得了很好的效果。

Compared with ordinary U-shaped and A-shaped Coriolis mass flowmeters, the micro-bend type Coriolis mass flowmeter has larger measurement error in measuring gas-liquid （air-water） two-phase flow, because it has higher natural frequency and smaller phase difference. Therefore, an experimental scheme is designed to perform the gas-liquid two-phase flow experiments using the Coriolis mass flow transmitter developed by our research group. The artificial neural network combining Back Propagation （BP） algorithm is utilized to conduct the modeling for the measurement error and obtain the error model; the online correction of the measurement error is realized in real time for the gas-liquid two-phase flow. The experiment results show that when the density drop varies from 0% to 30% , after online correction, the measurement error of the gas-liquid two-phase flow is reduced from the original maximum of -50% down to within -5% 3% , and good online correction results are achieved.