Optimal Feature Selection for Distributed Data-Driven Process Monitoring

Khatib, Shaaz; Daoutidis, Pródromos

doi:10.1021/acs.iecr.9b03872

Cited by 8 publications

(6 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Taylor bubbles' two‐phase flow characteristics with varying the location of a single injection nozzle from the centre of the column: (A) Taylor bubble rise velocity in a 240‐mm diameter column of 360 Pa · s viscosity silicone oil, ● from electrical capacitance tomography (ECT),  calculated from Equation (), C B c = 2.27; (B) experimental and analytical lengths of Taylor bubble, ● Taylor bubble (ECT),  Taylor bubble [ 32 ] ; (C) frequency of Taylor bubbles rising in the column; and (D) liquid film thickness around large bubbles. In (A–D) a single gas injection point at different positions is used at 0.04 m/s gas superficial velocity.…”

Section: Resultsmentioning

confidence: 99%

“…The no‐slip condition at the pipe wall may lead to a higher shear stress in the gas/liquid interface which could increase the chance of bubble coalescence, leading to increase in the length of the bubbles. Experimental bubble length was compared with the approach of Khatib and Richardson, [ 32 ] who proposed an equation to calculate the length of the Taylor bubble numerically. In their equation, they used the values of the probability density function (PDF) averaged void fraction to calculate the Taylor bubble and liquid slug length, L TB and L S , respectively.…”

Section: Resultsmentioning

confidence: 99%

“…In general, the analytical results calculated from the approach proposed by Khatib and Richardson [ 32 ] show no considerable change with moving the gas injection points.…”

Section: Resultsmentioning

confidence: 99%

See 2 more Smart Citations

An experimental study on the effect of gas injection configuration on flow characteristics in high viscosity oil columns

et al. 2021

View full text Add to dashboard Cite

Gas‐viscous liquid bubbly and slug flow are very common in petroleum, chemical, bioengineering, polymer, and food processing. However, there is a major knowledge gap in two‐phase flow research in the design of gas injectors/distributers in very high viscosity oil systems. The present study investigates the effect of gas injection methods in columns containing very high viscosity oils (i.e., realistic liquids), and more specifically using 360 Pa · s viscosity oil in a 240‐mm diameter column. The effects that the radial positioning, number of gas nozzles, and their distance from each other have on the structure of the flow in viscous liquids are presented in detail. Electrical capacitance tomography (ECT) is used to extract experimental data. Void fraction, bubble velocity, frequency, liquid film thickness, and bubble length were measured and analyzed at different radial gas injection positions. It has been observed that bubble length increases significantly by 0.3 m when the injection nozzle is located next to the wall of the pipe. Bubble velocity and length also increase by 0.217 m/s and 3.6 m, respectively, with increasing gas flowrate when multiple injection points are used. Increasing the distance between the gas injection points increased bubbles' length by 1.2 m. Bubbles' velocity and frequency (at higher gas flow rate) were also increased.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

An experimental study on the effect of gas injection configuration on flow characteristics in high viscosity oil columns

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The solid requirement of production safety and sustainable operation keeps motivating novel approaches for effectively monitoring the operating condition of modern industrial processes 1–4 . Nowadays, the great achievements in Industry 4.0 have been popularizing wider application of data‐driven process monitoring techniques 3–5 .…”

Section: Introductionmentioning

confidence: 99%

Just‐in‐time latent autoregressive residual generation for dynamic process monitoring

Hu,

Chang

2023

Journal of Chemometrics

View full text Add to dashboard Cite

With a goal of timely and adaptively exploiting the inconsistency inherited in the monitored samples of current interest, a novel dynamic process monitoring method based on just‐in‐time latent autoregressive residual generation (JITLAR2G) model is proposed. Different from the mainstream dynamic modeling and monitoring methods which usually train a signature generating mechanism and then repeatedly apply it for online monitored samples, the proposed JITLAR2G‐based approach provides a JITLAR2G model for the online monitored samples after data augmentation, so that the corresponding inconsistency within the given consecutive samples could be timely and adaptively uncovered. Instead of expressing the time‐serial relationship that generally accepted by the normal samples in the given dataset, solving the objective function designed for JITLAR2G in a just‐in‐time manner can adaptively and correspondingly seek but only one projecting vector as well as coefficient vector to generate residual, which points to the potential inconsistency inherited in the monitored samples, for the sole purpose of fault detection. As demonstrated through comparisons, the proposed JITLAR2G model can consistently guarantee its effectiveness, in terms of reducing both false alarm rate and missed alarm rate, for dynamic process monitoring, the salient performance achieved by the proposed JITLAR2G‐based method in contrast to the counterparts can be always confirmed.

show abstract

“…In Ref. [37], faulty information was taken advantage of, and process decomposition methods that can lead to minimal rate of missed detection of process faults were proposed. Optimal monitoring performance can be achieved by using these methods to partition a process into groups.…”

Section: Introductionmentioning

confidence: 99%

Community detection based process decomposition and distributed monitoring for large‐scale processes

et al. 2022

View full text Add to dashboard Cite

Distributed architectures wherein multiple decision-making units are employed to coordinate their decision-making/actions based on real-time communication have become increasingly important for monitoring processes that have large scales and complex structures. Typically, the development of a distributed monitoring scheme involves two key steps, that is, the decomposition of the process into subsystems, and the design of local monitors based on the configured subsystem models. In this article, we propose a distributed process monitoring approach that tackles both steps for large-scale processes. A data-driven process decomposition approach is proposed by leveraging community structure detection to divide variables into subsystems optimally via finding a maximal value of the metric of modularity. A two-layer distributed monitoring scheme is developed where local monitors are designed based on the configured subsystems of variables using canonical correlation analysis. Innersubsystem interactions and inter-subsystem interactions are tackled by the two layers separately, such that the sensitivity of this monitoring scheme to certain types of faults is improved. We utilize a numerical example to illustrate the effectiveness and superiority of the proposed method. It is then applied to a simulated wastewater treatment process.

show abstract

Optimal Feature Selection for Distributed Data-Driven Process Monitoring

Cited by 8 publications

References 43 publications

An experimental study on the effect of gas injection configuration on flow characteristics in high viscosity oil columns

An experimental study on the effect of gas injection configuration on flow characteristics in high viscosity oil columns

Just‐in‐time latent autoregressive residual generation for dynamic process monitoring

Community detection based process decomposition and distributed monitoring for large‐scale processes

Contact Info

Product

Resources

About