明渠效应在多相计量中的应用研究  

Experimental Studies of Flume-flow Model in Multiphase Flow Metering

作　　者：林军[1] 董守平[1] 

机构地区：[1]中国石油大学(北京)机电工程学院流体测试实验室,北京102249

出　　处：《实验力学》2005年第1期44-50,共7页Journal of Experimental Mechanics

摘　　要：在多相流体力学和多相流测量方面大量的基础理论研究和实验研究的基础上,提出了全新的明渠流模型等多个流量计测模型用于多相流量计量。通过自主开发的智能型多相流量计测系统在室内多相流测试环道上,对模型的运用效果做了大量的实验研究。并且利用基于神经网络技术和模糊模式识别技术开发的计测软件对实验数据进行处理,结果表明:液相计测数据中 80%以上误差位于±5%的范围以内,除个别小流量外,所有误差位于±10%的范围以内;气相计测结果中 90%以上误差位于±10%的范围以内, 97%的误差位于±15%的范围以内。实验数据表明:该计测模型可以适用于多相流中不同粘度的液相流量计量;可适用于较宽的气、液相流量变化范围,模型计测误差稳定在可接受的水平。A new multiphase flow metering model named flume-flow model was presented based on theoretical and experimental studies of multiphase fluid mechanics and multiphase flow testing. The experiment was carried on by the independent intelligent multiphase flow metering and testing system, and by the indoor multiphase flow loop. The data was processed by the special software developed based on neural network and fuzzy model identification technique. The effect of the flume-flow model was evaluated by a mass experimental data. Under the laboratory conditions, more than 80% of errors of the model measured liquid rates were within ±5% of the reference flow rates. Almost all the errors of the liquid rates were within ±10%, except a few of flow rate points. More than 90% of errors of the model measured gas rates were within ±10% of the reference flow rates, and 97% within ±15%. The results show that errors of the model are reasonable acceptable despite the variations of viscosity of liquid, and in a wide gas and liquid flow rate range.

关 键 词：多相计量 应用 效应 基础理论研究 多相流体力学 多相流量计量 模糊模式识别 神经网络技术 实验研究 计测模型 多相流测量 自主开发 运用效果 实验数据 技术开发 流量变化 误差 明渠流 智能型 行处理 液相 小流量 数据表 适用 

分 类 号：O235[理学—运筹学与控制论] TE863.1[理学—数学]