基于递归定量特征的气-液两相流流型融合识别  

Fusion Identification Method for Gas-Liquid Two-Phase Flow Regime Based on Recurrence Quantification Characteristics

作　　者：孙斌[1] 李超[1] 周云龙[1] 

机构地区：[1]东北电力大学能源与机械工程学院,吉林省吉林市132012

出　　处：《核动力工程》2009年第6期57-62,共6页Nuclear Power Engineering

基　　金：国家自然科学基金资助项目(50706006)

摘　　要：为了进一步提高流型识别的准确率,针对气-液两相流压差波动信号的非平稳特征,提出了一种基于递归定量分析(RQA)和多传感器数据融合技术的流型识别方法。该方法首先采用RQA方法提取压差波动信号的非线性特征参数,对3个不同取压间距压差波动信号的特征参数进行特征层融合,构成融合特征向量,并运用融合的特征向量对支持向量机进行训练并识别流型。对水平管内空气-水两相流4种典型流型的识别结果表明,经过多传感器数据融合,识别结果的可信度明显提高。To increase further the accuracy of flow regime and considering the non-stationary characteristics of differential pressure fluctuation signals of gas-liquid two-phase flow, the flow regime identification method based on recurrence quantification analysis （RQA） and multi-sensor data fusion techniques is put forward. First of all, the recurrence quantification analysis method is used to extract the nonlinear feature parameters of the differential pressure fluctuation signals of gas-liquid two-phase flow, and data fusion of feature layer is conducted by QRA feature parameters of differential pressure signals of three pressure measure intervals, and composes the fusion feature vectors. The fused characteristic vector are input into the support vector machine for identify flow regime. The identification results for four typical flow regimes of air-water two-phase flow in horizontal pipe has shown that the reliability of the identification result is improved evidently.

关 键 词：流型识别 递归定量分析 信息融合 支持向量机 

分 类 号：O359.1[理学—流体力学]