欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出一种基于截面电导信息的油水两相流相含率估计方法,在对电阻层析成像系统测试数据特征进行降维的基础上,利用径向基函数神经网络建立流型辨识模型,并对每一种流型采用基于样本矩阵非线性变换的非线性偏最小二乘（NLPLS）法建立相含率估计模型.动态实验结果表明,所得的相含率估计绝对误差低于5%,.将本方法和不分流型的单模型方法及传统偏最小二乘方法进行对比,证明所提出的相含率估计方法能实现更准确的估计.
英文摘要：
A method for estimating the phase holdup based on cross sectional conductance measurement was proposed. Based on the feature extraction of the original data,the flow regime identification model was established by using the radial basis function neural network. For each flow regime,the phase holdup estimation model was established by using nonlinear partial least squares（NLPLS）method which was based on the nonlinear transformation of the sample matrix. The estimated result error was less than 5%,. Compared with the single model method and the traditional partial least squares method,the proposed method is proved to achieve a more accurate estimation.
同期刊论文项目
同项目期刊论文
3D reconstruction of single rising bubble in water using digital image processing and characteristic
Investigation of upward cocurrent gas-liquid pipe flow using ultrafast X-ray tomography and wire-mes
期刊信息
位置：