摘要
在测量含气液体流量时,科氏质量流量计测量误差大,并且会出现流量管停振的现象。针对含气液体流量测量时的问题,设计了一套两相流实验装置及流程,研制了基于DSP的数字科氏质量流量变送器系统,进行气液两相流实验,通过改变液体流量点和含气量,采集实验数据样本;提出采用BP神经网络的方法,用于测量误差样本的建模与预测,并将网络训练后的模型参数植入DSP,实时实现对含气液体流量的在线测量与校正。实验结果表明,在含气量达到20%时,所研制的数字科氏质量流量变送器系统能有效维持流量管振动,经过校正方法修正后的实时测量误差优于±4%,大大减小了液体流量的测量误差。
There are large measurement errors in Coriolis mass flowmeter and the flowtube may stall when liquid flow mixed with gas is measured.Aiming at the problem in the flow measurement of liquid mixed with gas,an experimental rig and its operation process were designed,and a digital Coriolis mass flow transmitter based on DSP was developed to perform gas-liquid two-phase flow experiments.Ex-periment data were obtained through changing different liquid flow values and gas volume fractions,and Back Propagation (BP)neural network algorithm was used to carry out the modeling and forecast of the measurement errors,the model parameters of the network after training were implanted into a DSP to realize the online flow measurement and correction of liquid mixed with gas.The experiment results show that when the gas volume fraction reaches 20%,the developed Coriolis mass flow transmitter can maintain the flowtube oscillation;after error correction the metering error is better than ±4%,the measurement error of liquid flow is reduced significantly.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2014年第8期1796-1802,共7页
Chinese Journal of Scientific Instrument
基金
中航工业产学研合作创新工程(CXY2011HFGD23)资助
关键词
科氏质量流量计
两相流实验
BP神经网络
在线测量与校正
Coriolis mass flowmeter
two-phase flow experiment
BP neural network
online measurement and correction