基于奇异值分解和最小二乘支持向量机的气-液两相流流型识别方法  

Identification Method for Gas-Liquid Two-Phase Flow Regime Based on Singular Value Decomposition and Least Square Support Vector Machine

作　　者：孙斌[1] 周云龙[1] 赵鹏[1] 关跃波[1] 

机构地区：[1]东北电力大学能源与机械工程学院,吉林132012

出　　处：《核动力工程》2007年第6期62-66,共5页Nuclear Power Engineering

摘　　要：针对气-液两相流压差波动信号的非平稳特征和BP神经网络学习算法收敛速度慢、易陷入局部极小值等问题,提出了一种基于奇异值分解和最小二乘支持向量机(LS-SVM)的流型识别方法。该方法首先采用经验模态分解将气-液两相流压差波动信号分解为多个平稳的固有模态函数之和,并形成初始特征向量矩阵;对初始特征向量矩阵进行奇异值分解,得到矩阵的奇异值,将其作为流型的特征向量,根据LS-SVM分类器的输出结果来识别流型。对水平管内空气-水两相流4种典型流型进行识别,结果表明,与神经网络相比,该方法具有更高的识别率和识别速度。Aiming at the non-stationary characteristics of differential pressure fluctuation signals of gas-liquid two-phase flow, and the slow convergence of learning and liability of dropping into local minima for BP neural networks, flow regime identification method based on Singular Value Decomposition （SVD） and Least Square Support Vector Machine （LS-SVM） is presented. First of all, the Empirical Mode Decomposition （EMD） method is used to decompose the differential pressure fluctuation signals of gas-liquid two-phase flow into a number of stationary Intrinsic Mode Functions （IMFs） components from which the initial feature vector matrix is formed. By applying the singular value decomposition technique to the initial feature vector matrixes, the singular values are obtained. Finally, the singular values serve as the flow regime characteristic vector to be LS-SVM classifier and flow regimes are identified by the output of the classifier. The identification result of four typical flow regimes of air-water two-phase flow in horizontal pipe has shown that this method achieves a higher identification rate.

关 键 词：流型识别 经验模态分解 奇异值分解 最小二乘支持向量机 

分 类 号：O359.1[理学—流体力学]