A Novel Method for Gas-Liquid Flow Pattern Recognition Based on PDF and SVM

2019 ◽  
Author(s):  
Dong-Xu Wang ◽  
Qi-Hui Hu ◽  
Yu-Xing Li ◽  
Quan Wang ◽  
Shuang Li
Keyword(s):  
Pattern Recognition ◽  
Flow Pattern ◽  
Slug Flow ◽  
Support Vector ◽  
Liquid Holdup ◽  
Horizontal Pipes ◽  
Pattern Recognition Methods ◽  
Novel Method ◽  
Gas Liquid Flow ◽  
Probability Density Function Pdf

Abstract Currently, most of the traditional flow pattern recognition methods are based on the features of differential pressure signal, and the verification data is majorly derived from the horizontal or almost horizontal pipes. This paper proposes a new method that combines the probability density function (PDF) of the liquid holdup signal and the support vector machine (SVM). Results demonstrate the capability of the proposed algorithm to effectively recognize the stratified flow, bubble flow, slug flow, and severe slug flow and eliminate subjective factors.

GAS-LIQUID FLOW PATTERN RECOGNITION USING DIMENSIONLESS NUMBERS

2019 ◽  
Author(s):  
Caio Araujo ◽  
Tiago Ferreira Souza ◽  
Maurício Figueiredo ◽  
valdir estevam ◽  
Ana Maria Frattini Fileti
Keyword(s):  
Pattern Recognition ◽  
Flow Pattern ◽  
Liquid Flow ◽  
Dimensionless Numbers ◽  
Gas Liquid Flow

GAS-LIQUID FLOW-PATTERN TRANSITION IN HORIZONTAL PIPES – ANALYSIS OF REPORTED MODELS

2018 ◽  
Author(s):  
André Mendes Quintino ◽  
Davi Lotfi Lavor Navarro da Rocha ◽  
Oscar Mauricio Hernandez Rodriguez
Keyword(s):  
Flow Pattern ◽  
Liquid Flow ◽  
Horizontal Pipes ◽  
Flow Pattern Transition ◽  
Gas Liquid Flow

scholarly journals Optimization and Extended Applicability of Simplified Slug Flow Model for Liquid-Gas Flow in Horizontal and Near Horizontal Pipes

Energies ◽  
2020 ◽  
Vol 13 (4) ◽  
pp. 842
Author(s):  
Tea-Woo Kim ◽  
Nam-Sub Woo ◽  
Sang-Mok Han ◽  
Young-Ju Kim
Keyword(s):  
Slug Flow ◽  
Pressure Loss ◽  
Gas Flow ◽  
Flow Model ◽  
Flow Coefficient ◽  
Coefficient Of Determination ◽  
Liquid Holdup ◽  
Two Phase ◽  
Horizontal Pipes ◽  
Image Velocimetry

The accurate prediction of pressure loss for two-phase slug flow in pipes with a simple and powerful methodology has been desired. The calculation of pressure loss has generally been performed by complicated mechanistic models, most of which require the iteration of many variables. The objective of this study is to optimize the previously proposed simplified slug flow model for horizontal pipes, extending the applicability to turbulent flow conditions, i.e., high mixture Reynolds number and near horizontal pipes. The velocity field previously measured by particle image velocimetry further supports the suggested slug flow model which neglects the pressure loss in the liquid film region. A suitable prediction of slug characteristics such as slug liquid holdup and translational velocity (or flow coefficient) is required to advance the accuracy of calculated pressure loss. Therefore, the proper correlations of slug liquid holdup, flow coefficient, and friction factor are identified and utilized to calculate the pressure gradient for horizontal and near horizontal pipes. The optimized model presents a fair agreement with 2191 existing experimental data (0.001 ≤ μL ≤ 0.995 Pa∙s, 7 ≤ ReM ≤ 227,007 and −9 ≤ θ ≤ 9), showing −3% and 0.991 as values of the average relative error and the coefficient of determination, respectively.

Computational Estimation of Bubble Gas Velocity for Slug Flow Pattern in Horizontal Pipes Using Image Processing Technique

2012 ◽  
Author(s):  
María T. Valecillos ◽  
Carlos H. Romero ◽  
María A. Márquez ◽  
Sissi D. Vergara
Keyword(s):  
Image Processing ◽  
Flow Pattern ◽  
Slug Flow ◽  
High Speed ◽  
Processing Technique ◽  
Image Processing Technique ◽  
Gas Velocity ◽  
Two Phase ◽  
Horizontal Pipes ◽  
Computational Estimation

Two-phase slug flow pattern is one of the most common flow patterns present in many industries, therefore its study becomes relevant. The aim of this work was to develop an automated computational program to determine the bubble gas velocity associated to gas-liquid two-phase slug flow by using video digital image processing technique. In order to obtain the images for the analysis, experiments were carried out using a pipe bench for air-water two-phase flow. The experimental facility is located in Simon Bolivar University, in Venezuela. The system has three pipes with different internal diameters and can be rotated around its axis and fixed at any inclination angle from horizontal to vertical flow. The tests were run in a horizontal pipeline of 0.03175m of internal pipe diameter and 8m long. For slug flow visualization a high speed camera Kodak Ektapro 4540mx imager was used. The camera was located in an x/D relation corresponding to 249 from the pipe inlet, ensuring the complete development of the flow. The camera allowed a maximum acquisition velocity of 4500 frames per second. The superficial velocity range was 0.16–1.79m/s and 0.16–1.26m/s for air and water, respectively. To summarize, 165 tests were performed and 1320000 images were analyzed with 20 flow rate combinations. The computational application was validated by comparing it with the velocities measured manually over selected images. Results obtained were compared to several correlations such as Bendiksen [1], Cook & Behnia [2] and Wang et al. [3].

scholarly journals Electronic Nose-Based Technique for Rapid Detection and Recognition of Moldy Apples

2019 ◽  
Author(s):  
Wenshen Jia ◽  
Gang Liang ◽  
Hui Tian ◽  
Jing Sun ◽  
Cihui Wan
Keyword(s):  
Neural Network ◽  
Pattern Recognition ◽  
Electronic Nose ◽  
Gas Sensors ◽  
Penicillium Expansum ◽  
Support Vector ◽  
Linear Discriminant ◽  
Pattern Recognition Methods ◽  
Group A ◽  
Group B

In this paper, PEN3 electronic nose was used to detect and recognize fresh and moldy apples (inoculated with Penicillium expansum and Aspergillusniger) taken Golden Delicious apples as model subject. Firstly, the apples were divided into two groups: apples only inoculated with different molds (Group A) and mixed apples of inoculated apples with fresh apples (Group B). Then the characteristic gas sensors of the PEN3 electronic nose that were most closely correlated with the flavor information of the moldy apples were optimized and determined, which can simplify the analysis process and improve the accuracy of results. Four pattern recognition methods, including linear discriminant analysis (LDA), backpropagation neural network (BPNN), support vector machines (SVM) and radial basis function neural network (RBFNN), were then applied to analyze the data obtained from the characteristic sensors, respectively, aiming at establishing the prediction model of flavor information and fresh/moldy apples. The results showed that only the gas sensors of W1S, W2S, W5S, W1W and W2W in the PEN3 electronic nose exhibited strong signal response to the flavor information, indicating were most closely correlated with the characteristic flavor of apples and thus the data obtained from these characteristic sensors was used for modeling. The results of the four pattern recognition methods showed that BPNN presented the best prediction performance for the training and validation sets for both the Group A and Group B, with prediction accuracies of 96.29% and 90.00% (Group A), 77.70% and 72.00% (Group B), respectively. Therefore, it first demonstrated that PEN3 electronic nose can not only effectively detect and recognize the fresh and moldy apples, but also can distinguish apples inoculated with different molds.

scholarly journals Gas–liquid Flow Pattern Recognition Based on Wavelet Packet Energy Entropy of Vortex-induced Pressure Fluctuation

2013 ◽  
Vol 13 (2) ◽  
pp. 83-88 ◽  
Author(s):  
Zhiqiang Sun ◽  
Shuai Shao ◽  
Hui Gong
Keyword(s):  
Pattern Recognition ◽  
Flow Pattern ◽  
Liquid Flow ◽  
Pressure Fluctuation ◽  
Bluff Body ◽  
Flow Direction ◽  
Wavelet Packet ◽  
Flow Patterns ◽  
Energy Entropy ◽  
Gas Liquid Flow

Here we report a novel flow-pattern map to distinguish the gas-liquid flow patterns in horizontal pipes at ambient temperature and atmospheric pressure. The map is constructed using the coordinate system of wavelet packet energy entropy versus total mass flow rate. The wavelet packet energy entropy is obtained from the coefficients of vortex-induced pressure fluctuation decomposed by the wavelet packet transform. A triangular bluff body perpendicular to the flow direction is employed to generate the pressure fluctuation. Experimental tests confirm the suitability of the wavelet packet energy entropy as an ideal indicator of the gas-liquid flow patterns. The overall identification rate of the map is 92.86%, which can satisfy most engineering applications. This method provides a simple, practical, and robust solution to the problem of gas-liquid flow pattern recognition.

Surfactant Use for Slug Flow Pattern Suppression in a Horizontal Pipe

2003 ◽  
Author(s):  
R. J. Wilkens ◽  
S. R. Glassmeyer ◽  
G. J. Rosebrock ◽  
K. M. Storage ◽  
T. M. Storage
Keyword(s):  
Flow Pattern ◽  
Liquid Flow ◽  
Pipe Flow ◽  
Slug Flow ◽  
Flow Patterns ◽  
Flow Transition ◽  
Horizontal Pipe ◽  
Surfactant Additive ◽  
Gas Liquid Flow

A set of experiments was performed to study flow pattern suppression in gas-liquid pipe flow by means of surfactant additive. Results suggest that addition of the surfactant to gas-liquid flow significantly reduces the occurrence of slug flow. In addition, previously unreported flow patterns were observed to exist between slug and dispersed bubble flows. It is concluded that new mechanisms for slug flow transition need to be considered.

A hybrid of fast K-nearest neighbor and improved directed acyclic graph support vector machine for large-scale supersonic inlet flow pattern recognition

2021 ◽  
pp. 095441002110086
Author(s):  
Huan Wu ◽  
Yong-Ping Zhao ◽  
Tan Hui-Jun
Keyword(s):  
Pattern Recognition ◽  
Flow Pattern ◽  
Directed Acyclic Graph ◽  
Large Scale ◽  
Computational Cost ◽  
Classification Error ◽  
Support Vector ◽  
Training Set ◽  
Inlet Flow ◽  
Supersonic Inlet

Inlet flow pattern recognition is one of the most crucial issues and also the foundation of protection control for supersonic air-breathing propulsion systems. This article proposes a hybrid algorithm of fast K-nearest neighbors (F-KNN) and improved directed acyclic graph support vector machine (I-DAGSVM) to solve this issue based on a large amount of experimental data. The basic idea behind the proposed algorithm is combining F-KNN and I-DAGSVM together to reduce the classification error and computational cost when dealing with big data. The proposed algorithm first finds a small set of nearest samples from the training set quickly by F-KNN and then trains a local I-DAGSVM classifier based on these nearest samples. Compared with standard KNN which needs to compare each test sample with the entire training set, F-KNN uses an efficient index-based strategy to quickly find nearest samples, but there also exists misclassification when the number of nearest samples belonging to different classes is the same. To cope with this, I-DAGSVM is adopted, and its tree structure is improved by a measure of class separability to overcome the sequential randomization in classifier generation and to reduce the classification error. In addition, the proposed algorithm compensates for the expensive computational cost of I-DAGSVM because it only needs to train a local classifier based on a small number of samples found by F-KNN instead of all training samples. With all these strategies, the proposed algorithm combines the advantages of both F-KNN and I-DAGSVM and can be applied to the issue of large-scale supersonic inlet flow pattern recognition. The experimental results demonstrate the effectiveness of the proposed algorithm in terms of classification accuracy and test time.

Sign in / Sign up

Export Citation Format

Share Document