潘慧, 李海广, 吴晅. 气液两相流压差波动信号的混沌特性及Volterra自适应短期预测研究[J]. 实验流体力学, 2020, 34(4): 102-108. DOI: 10.11729/syltlx20190077
引用本文: 潘慧, 李海广, 吴晅. 气液两相流压差波动信号的混沌特性及Volterra自适应短期预测研究[J]. 实验流体力学, 2020, 34(4): 102-108. DOI: 10.11729/syltlx20190077
PAN Hui, LI Haiguang, WU Xuan. A study on chaotic characteristics and short-term prediction of pressure difference fluctuation signal of gas-liquid two-phase flow in small channel[J]. Journal of Experiments in Fluid Mechanics, 2020, 34(4): 102-108. DOI: 10.11729/syltlx20190077
Citation: PAN Hui, LI Haiguang, WU Xuan. A study on chaotic characteristics and short-term prediction of pressure difference fluctuation signal of gas-liquid two-phase flow in small channel[J]. Journal of Experiments in Fluid Mechanics, 2020, 34(4): 102-108. DOI: 10.11729/syltlx20190077

气液两相流压差波动信号的混沌特性及Volterra自适应短期预测研究

A study on chaotic characteristics and short-term prediction of pressure difference fluctuation signal of gas-liquid two-phase flow in small channel

  • 摘要: 以空气和水为工质,在直径3.0 mm的水平圆管通道内进行压差波动信号实验研究。根据压差波动信号图及高速工业相机拍摄的流型图,结合相空间重构、Lyapunov指数判别法对压差波动信号进行混沌动力学分析,在其基础上对压差波动信号进行Volterra自适应短期预测。结果表明:混沌分析得到的吸引子图可以更准确地展现管道内气液两相流的流动特性;Volterra自适应短期预测模型可以有效地对管道内气液两相流的压差时间序列进行短期预测,对环状流、层状流、间歇流、段塞流的压差时间序列预测的相对误差分别为1.86%、0.71%、3.90%、2.49%。

     

    Abstract: Experimental research on the pressure difference fluctuation signal was conducted in the channel of a horizontal circular pipe with a diameter of 3.0 mm, using air and water as the working medium. According to the pressure difference fluctuation signal diagram and the flow pattern diagram taken by the high-speed camera, the chaotic dynamic analysis of the pressure difference fluctuation signal was carried out by using the phase space reconstruction and Lyapunov index discrimination methods. Then the Volterra adaptive short-term prediction of the pressure difference fluctuation signal was carried out. The experimental results show that the attractor diagram obtained by the chaos analysis can show the flow characteristics more accurately. The Volterra adaptive prediction model has the relative errors of 1.86%, 0.71%, 3.90% and 2.49% for the prediction of the pressure difference time series of the annular flow, the layered flow, the intermittent flow and the slug flow respectively, which can effectively make short-term prediction of pressure difference time series of the gas-liquid two-phase flow in the pipeline.

     

/

返回文章
返回