Artificial neural network approach for flow regime classification in gas–liquid–fiber flows based on frequency domain analysis of pressure signals

Xie, Tingli; Ghiaasiaan, S. Mostafa; Karrila, Seppo

doi:10.1016/j.ces.2004.02.017

Cited by 96 publications

(25 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The power spectral density is a frequency domain characteristic of a time series and is appropriate for the detection of frequency composition in a stochastic process (Matsumoto and Suzuki 1984). Assuming the process to be stationary and ergodic, the power spectral density function P x (f ) of a discrete-time signal x(n) is defined as the Fourier transform of the autocorrelation sequence R x (k) (Xie et al 2004):…”

Section: Power Spectral Analysismentioning

confidence: 99%

Neural networks approach for prediction of gas–liquid two-phase flow pattern based on frequency domain analysis of vortex flowmeter signals

Sun

Zhang

2007

Meas. Sci. Technol.

View full text Add to dashboard Cite

The identification of flow pattern is a basic and important issue in multiphase systems. Because of the complexity of phase interaction in gas-liquid two-phase flow, it is difficult to discern its flow pattern objectively. In this paper, a three-layer, feed-forward neural network was designed, and adopted inputs all representing the characteristics of the power spectral density distributions of dynamic differential pressure fluctuations were obtained from a vortex flowmeter. The validity of the adopted inputs for the flow pattern identification was evaluated by a proposed effectiveness factor. Results show that the designed neural networks predict the flow patterns successfully comparing with the flow pattern by visual observation. These findings provide the possibility of using only a vortex flowmeter for the identification of gas-liquid two-phase flow patterns with the help of neural networks.

show abstract

Section: Power Spectral Analysismentioning

confidence: 99%

Neural networks approach for prediction of gas–liquid two-phase flow pattern based on frequency domain analysis of vortex flowmeter signals

Sun

Zhang

2007

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…The measurement precision has continuously raised and its variety differed concerning the liquid or gas. In the beginning, most of them interfered with changing of the process features [160,148] e.g. mechanical, manometric, impulse flow-meters or rotameters.…”

Section: Two-phase Flows Non-invasive Measurements Techniquesmentioning

confidence: 99%

Computer methods for non-invasive measurement and control of two-phase flows: a review study

Wajman

2019

ITC

View full text Add to dashboard Cite

Nowadays, the application of advanced technologies in modern production systems is the main trend of development and technological progress in many industrial sectors. It is due to the still growing trends of energy-saving and production quality enhancement. Wherever in the production process the phase mixture is transported and it is not optimal or not economic, there is a need to develop a system which would be able to prevent any construction disaster, unexpected production line stopping or situation where for reasons of bad flow parameters, the final product is defective. This paper studies various sensors, measurement techniques and computer methods for signal processing and analysis to diagnose and control two-phase flows. Due to the possibility that the invasive measurement disturbs the process and changes it parameters and behavior especially in the location just after the measurement point and simultaneously does not provide any information about these changes it is unreliable for the diagnosis or control. Therefore, the non-invasive techniques commonly used for measurement of flows parameters are described. Depending on the industrial demands many of applications examples for non-invasive two-phase flows measurement and monitoring are given. This description for identifying the flow parameters is divided into features categories of this phenomena as void fraction distribution, velocity profile and flow regime. However, from these methods the high accuracy and short processing time is expected. The continued observation and monitoring of the process abnormalities can provide valuable information about its dynamic state and allow for real-time and automatic monitoring and control. Therefore, the development of advanced process control is one of the most important challenges to keep the flow regime on the given level and for instant and long-term energy saving, quality improvement.

show abstract

“…Based on past work involving vertical flows and other instrumentation, there was reason to believe that the use of neural network input values derived from time averaging or probability density functions have potential for reasonable success rates [13,14]. Thus, the seventh column lists the success 12 rates for flow pattern identification that used and , together as input for each experiment. Similarly, columns 8 through 10 list the success rates for input comprised of and with for each experimental run.…”

Section: Neural Network Studiesmentioning

confidence: 99%

“…Learning is achieved by the adjustment of these weights [12]. Past studies suggest that there is potential for using a neural network to objectively classify liquid-vapor distribution data [13][14][15].…”

Section: Introductionmentioning

confidence: 99%

Flow pattern identification of horizontal two-phase refrigerant flow using neural networks

Roman

Kreitzer

Ervin

et al. 2016

International Communications in Heat and Mass Transfer

View full text Add to dashboard Cite

In this work, electrical capacitance tomography (ECT) and neural networks were used to automatically identify two-phase flow patterns for refrigerant R-134a flowing in a horizontal tube. In laboratory experiments, high-speed images were recorded for human classification of liquid-vapor flow patterns. The corresponding permittivity data obtained from tomograms was then used to train feedforward neural networks to recognize flow patterns. An objective was to determine which subsets of data derived from tomograms could be used as input data by a neural network to classify nine liquidvapor flow patterns. Another objective was to determine which subsets of input data provide high identification success when analyzed by a neural network. Transitional flow patterns associated with common horizontal flow patterns were considered. A unique feature of the current work was the use of the vertical center of mass coordinate in pattern classification. The highest classification success rates occurred using neural network input which included the probability density functions (in time) for both spatially averaged permittivity and center of mass location in addition to the four statistical moments (in time) for spatially averaged permittivity data. The combination of these input data resulted in an average success rate of 98.1% for nine flow patterns. In addition, 99% of the experimental runs were either correctly classified or misclassified by only one flow pattern. NomenclatureA i Individual pixel area [m 2 ] A T Sum of all pixel areas [m 2 ] D Tube diameter [m] D b Bubble diameter [m] ECT Electrical capacitance tomography L Length [m] ____________________

show abstract

Artificial neural network approach for flow regime classification in gas–liquid–fiber flows based on frequency domain analysis of pressure signals

Cited by 96 publications

References 29 publications

Neural networks approach for prediction of gas–liquid two-phase flow pattern based on frequency domain analysis of vortex flowmeter signals

Neural networks approach for prediction of gas–liquid two-phase flow pattern based on frequency domain analysis of vortex flowmeter signals

Computer methods for non-invasive measurement and control of two-phase flows: a review study

Flow pattern identification of horizontal two-phase refrigerant flow using neural networks

Contact Info

Product

Resources

About