摘要
石油开采过程中油水流型对压力等参数影响很大,准确识别流型可以提高传输效率、降低开采成本。利用INV306型智能信号采集处理系统和电导探针测量系统,采集垂直上升管中三种不同油水两相流流型的电导波动信号。应用经验模态分解(EMD)对电导波动信号进行了IMF特征参数的提取,然后分别提取各模态的能量特征,并将其作为Elman神经网络的输入,训练并建立网络。用检测样本进行检测,结果表明:该方法可以准确识别出三种典型流型,具有较好的识别效果。
Oil-water flow patterns have a great impact on pressure and other parameters when being exploited, accurate identification of flow patterns can improve transmission efficiency and reducing production cost.Using the intelligent Signal Acquisition and Processing System type of INV306 and conductivity probe measurement system,the conductance fluctuation signals of three typical oil-water two-phase flow patterns in vertical upward pipe are collected.IMF characteristic parameters of conductance signals were extracted by applying of Empirical Mode Decomposition（EMD）,then energy characteristics of all modes were extracted,and they are trained as Elman neural network input in order to establish the network.With test samples,and the results show that： the method can accurately identify the three typical flow patterns with better recognition effect.
出处
《东北电力大学学报》
2011年第1期11-16,共6页
Journal of Northeast Electric Power University
关键词
油水两相流
经验模式分解
ELMAN神经网络
能量特征
Oil-water two phase flow
Empirical mode decomposition
Elman neural network
Energy characteristics
