Artificial neural network for prediction of SO2 removal and volumetric mass transfer coefficient in spray tower

Valera, Vinícius Y.; Codolo, Milene Costa; Martins, Tiago Dias

doi:10.1016/j.cherd.2021.03.008

Cited by 13 publications

(6 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The network was applied with the Levenberg–Marquardt backpropagation algorithm to train the overall mass transfer coefficient and then predict the mass transfer rate through the membrane module. Valera et al [ 19 ] and Hagan et al [ 20 ] found that Levenberg–Marquardt is the most common algorithm because it showed much more efficiency than others when the network contained no more than a few hundred weights. The activation function in the hidden layer was tangent sigmoid, given by the equation

where z is the weighted sum of the input.…”

Section: An Ann For Prediction Of the Overall Mass Transfer Coefficientmentioning

confidence: 99%

A Correlation of Overall Mass Transfer Coefficient of Water Transport in a Hollow-Fiber Membrane Module via an Artificial Neural Network Approach

et al. 2022

View full text Add to dashboard Cite

Water transport in a hollow-fiber membrane depends on mass convection around the tube, mass convection inside the tube, and water diffusion through the membrane tube. The performance of water transport is then explained by the overall mass transfer coefficient in hollow-fiber membranes. This study presents the prediction of overall mass transfer coefficients of water transport in a hollow-fiber membrane module by an artificial neural network (ANN) that is used for a humidifier of a vehicular fuel cell system. The input variables of ANN are collected from water transport experiments of the hollow-fiber membrane module that is composed of inlet flow rates, inlet relative humidity, system pressures, and operating temperatures. The experimental mass transfer coefficients are the targets of the training model, which are determined via the effectiveness analysis. When unknown data are applied to the ANN model, the correlation of the overall mass transfer coefficient predicts precise results with R = 0.99 (correlation coefficient). The ANN model shows good prediction capability of water transport in membrane humidifiers.

show abstract

where z is the weighted sum of the input.…”

Section: An Ann For Prediction Of the Overall Mass Transfer Coefficientmentioning

confidence: 99%

A Correlation of Overall Mass Transfer Coefficient of Water Transport in a Hollow-Fiber Membrane Module via an Artificial Neural Network Approach

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The other type of methods is based on machine learning method, such as tree model [13][14][15] , neural network model [16,17] and so on. In particular, Wang et al designed a network to automatically extract feature information from data and proposed a deep learning framework for long-term time sequence forecasting based on the extracted features [18] . Yang et al combined different tree-based machine learning algorithms into an enhanced machine learning model and predicted bubble point pressure based on composition data and temperature [19] .…”

Section: Introductionmentioning

confidence: 99%

“…(t), the proposed method is composed of the following steps: in(18) and the current trend sequence () cc sM in(19), respectively.…”

mentioning

confidence: 99%

Alarm state forecasting from similar trends in historical time sequences based on Bayesian estimators

Kang

et al. 2023

Second International Conference on Advanced Manufacturing Technology and Manufacturing Systems (ICAMTMS 2023)

View full text Add to dashboard Cite

Predicting the upcoming alarm state helps industrial plant operators take timely actions to avoid the negative effects of abnormal conditions. This paper proposes a new data-driven method for trend forecasting in a process variable. The proposed method has two important features: (i) alarm states are predicted in real-time whenever a new trend occurs in the current time sequence, which helps industrial operators take timely actions at the early stage of abnormal conditions; (ii) alarm states are predicted with quantitative probabilities and credibility intervals. By contrast, existing early warning methods are either suitable for specific scenarios where a short prediction step length is needed or cannot provide a quantitative measurement of prediction results. The proposed method firstly captures qualitative trends in the current and historical time sequences and clusters these trends into several categories. Secondly, current and historical trend sequences are obtained according to the cluster results. Finally, recent trends in the current trend sequence are chosen, and alarm states are predicted based on the historical trend sequence having the same sub-sequence. A Bayesian estimator is exploited to calculate the posterior probabilities and their credibility intervals for predicted alarm states. The optimal predicted alarm state is the one with the largest lower bound of posterior probability as a quantitative measurement of prediction accuracy. The effectiveness of the proposed method is illustrated via numerical examples based on a well-accepted benchmark of the Tennessee Eastman process.

show abstract

“…22,23 The spray tower is the core equipment in the wet flue-gas desulfurization process of a coal-fired power plant boiler tail gas. [24][25][26] In the spray tower, the limestone slurry washes the flue gas, essentially a gas-liquid two-phase countercurrent contact process, and a gas-liquid mass transfer is accompanied by a two-phase flow process. During the gas-liquid two-phase countercurrent contact process, drops will continue to leave the spray layer under the action of the flue gas and enter the mist eliminator for gas-liquid separation.…”

Section: Introductionmentioning

confidence: 99%

“…Although the application of this technology has greatly reduced the concentration of exhaust smoke pollutants, the entrainment of drops also leads to a series of new problems, such as gypsum rain and acid rain that pollute the power plant and surrounding environment, while also corroding equipment 22,23 . The spray tower is the core equipment in the wet flue‐gas desulfurization process of a coal‐fired power plant boiler tail gas 24–26 . In the spray tower, the limestone slurry washes the flue gas, essentially a gas–liquid two‐phase countercurrent contact process, and a gas–liquid mass transfer is accompanied by a two‐phase flow process.…”

Section: Introductionmentioning

confidence: 99%

Drop breakup and entrainment in the updraft

Liu

Wang

et al. 2022

AIChE Journal

View full text Add to dashboard Cite

This study aims to investigate the characteristics of gas-liquid countercurrent contact processes. In spray towers or other applications, several drops containing pollutants are entrained by the updraft flue gas, which can easily cause environmental pollution. Traditionally, this drop entrainment phenomenon is alleviated by increasing the diameter of the drops. However, the breakup of a large drop would also cause drop entrainment to become serious, a process referred to as secondary atomization. Herein, we propose the boundary of three drop modes in the updraft: drop falling mode, reverse entrainment mode, and breakup entrainment mode. The critical Weber number (We) is the key dimensionless number marking the beginning of the drop breakup. The ratio of the drag force to gravity and We are proposed as criteria for the drop entrainment.

show abstract

Artificial neural network for prediction of SO2 removal and volumetric mass transfer coefficient in spray tower

Cited by 13 publications

References 37 publications

A Correlation of Overall Mass Transfer Coefficient of Water Transport in a Hollow-Fiber Membrane Module via an Artificial Neural Network Approach

A Correlation of Overall Mass Transfer Coefficient of Water Transport in a Hollow-Fiber Membrane Module via an Artificial Neural Network Approach

Alarm state forecasting from similar trends in historical time sequences based on Bayesian estimators

Drop breakup and entrainment in the updraft

Contact Info

Product

Resources

About