The article is an attempt to compile the results of CFD liquid flow simulation through pipeline section containing hydraulic elbow with the results of ultrasonic flow measurements. To carry out the measurements behind...The article is an attempt to compile the results of CFD liquid flow simulation through pipeline section containing hydraulic elbow with the results of ultrasonic flow measurements. To carry out the measurements behind the flow disturbing element(hydraulic elbow), an ultrasonic flowmeter with applied head set in accordance with the Z-type system was used. For comparative purposes, a flow simulation for 3 different turbulence models(k-epsilon, SST and SSG) was performed. It was found that with a proper ultrasonic flowmeter heads configurations, it is possible to measure the flow rate disturbed by the hydraulic elbow at any distance from the source of the disturbance. It has to use appropriate correction factor that can be determined by knowing the flow velocity profile equation. Based on comparison of CFD simulation results with experimental data, the accuracy/purposefulness of using individual turbulence models in the case of discussed hydraulic installation was evaluated.展开更多
气液两相流参数测量的准确性和实时性对火力发电、化工、石油等工业生产具有重要意义。为提升气液两相流液相流量测量速度和精度,通过对电阻层析成像(electrical resistance tomography,ERT)技术和长短时记忆(long and short time memor...气液两相流参数测量的准确性和实时性对火力发电、化工、石油等工业生产具有重要意义。为提升气液两相流液相流量测量速度和精度,通过对电阻层析成像(electrical resistance tomography,ERT)技术和长短时记忆(long and short time memory,LSTM)神经网络的研究,提出了一种基于LSTM神经网络的气液两相流流量测量方法。利用16电极ERT系统分别采集了弹状流、泡状流两种流型对应的阵列电阻值,并将其作为LSTM神经网络的输入,电磁流量计采集的液相流量值作为理想输出,通过网络训练得到了2种流型下的液相流量测量模型。仿真实验结果表明:利用该方法在弹状流和泡状流下的均方根误差分别为5.359 m^(3)/h和4.088 m^(3)/h,平均引用误差低于3%,测量精度和速度优于BP神经网络,能够实现对两种流型下液相体积流量的准确测量。展开更多
文摘The article is an attempt to compile the results of CFD liquid flow simulation through pipeline section containing hydraulic elbow with the results of ultrasonic flow measurements. To carry out the measurements behind the flow disturbing element(hydraulic elbow), an ultrasonic flowmeter with applied head set in accordance with the Z-type system was used. For comparative purposes, a flow simulation for 3 different turbulence models(k-epsilon, SST and SSG) was performed. It was found that with a proper ultrasonic flowmeter heads configurations, it is possible to measure the flow rate disturbed by the hydraulic elbow at any distance from the source of the disturbance. It has to use appropriate correction factor that can be determined by knowing the flow velocity profile equation. Based on comparison of CFD simulation results with experimental data, the accuracy/purposefulness of using individual turbulence models in the case of discussed hydraulic installation was evaluated.
文摘气液两相流参数测量的准确性和实时性对火力发电、化工、石油等工业生产具有重要意义。为提升气液两相流液相流量测量速度和精度,通过对电阻层析成像(electrical resistance tomography,ERT)技术和长短时记忆(long and short time memory,LSTM)神经网络的研究,提出了一种基于LSTM神经网络的气液两相流流量测量方法。利用16电极ERT系统分别采集了弹状流、泡状流两种流型对应的阵列电阻值,并将其作为LSTM神经网络的输入,电磁流量计采集的液相流量值作为理想输出,通过网络训练得到了2种流型下的液相流量测量模型。仿真实验结果表明:利用该方法在弹状流和泡状流下的均方根误差分别为5.359 m^(3)/h和4.088 m^(3)/h,平均引用误差低于3%,测量精度和速度优于BP神经网络,能够实现对两种流型下液相体积流量的准确测量。
