Intelligent Steam Power Plant Boiler Waterwall Tube Leakage Detection via Machine Learning-Based Optimal Sensor Selection

Khalid, Salman; Lim, Woo Cheol; Kim, Heung Soo; Oh, Yeongtak; Youn, Byeng D.; Kim, Hee‐Soo; Bae, Yong-Chae

doi:10.3390/s20216356

Cited by 17 publications

(5 citation statements)

References 48 publications

(52 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Jiheon Kang et al proposed a water leakage detection and localization system utilizing a one-dimensional convolutional neural network (CNN) in conjunction with SVM (1D-CNN-SVM) improving the performance using a graph-based local search algorithm [22]. Salman Khalid et al adopt machine learning to support the sensor selection in waterwall tube leakage detection for steam power plants [23].…”

Section: State Of the Artmentioning

confidence: 99%

See 1 more Smart Citation

A Machine Vision—Based Pipe Leakage Detection System for Automated Power Plant Maintenance

Bao

Fan

et al. 2022

Sensors

View full text Add to dashboard Cite

Industrial pipework maintenance inspection can be automated through machine vision-based effusion monitoring. However, colorless effusions such as water can be difficult to detect in a complex industrial environment due to weak illumination and poor visibility of the background. This paper deploys the reflective characteristics of effusion and its lower temperature compared to the environment in order to develop an automatic inspection system for power plant pipeworks’ maintenance. Such a system is aimed at detecting the colorless fluid effusion based on dual source images and a contour features algorithm. In this respect, a visible light source unit highlights the reflective features of the effusion edge. Meanwhile, high-definition images of the potential effusion are acquired under both visible and infrared lights. A customized image processing procedure extracts the potential effusion features from the infrared image to retrieve the region of interest for segmentation purposes and transfer such information to the visible light image to determine the effusion contour. Finally, a decision-making support tool based on the image contour closure is enabled for classification purposes. The implementation of the proposed system is tested on a real industrial environment. Experimental results show a classification accuracy up to 99%, demonstrating excellent suitability in meeting industrial requirements.

show abstract

Section: State Of the Artmentioning

confidence: 99%

“…Waterwall boiler tube leakage is the most frequent cause of failure in thermal power plants [23]. One specific study has shown a strong correlation between internal cracks and leakage phenomena, with the development of cavity nucleation and growth on grain boundaries eventually yielding macro-intergranular cracks [33].…”

Section: State Of the Artmentioning

confidence: 99%

A Machine Vision—Based Pipe Leakage Detection System for Automated Power Plant Maintenance

Bao

Fan

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…With the development of technology such as data mining and machine learning technologies, fault detection methods of industrial boilers have changed from knowledge‐based monitoring methods to data‐driven monitoring methods. [ 14 ] The neural networks, [ 15–17 ] principal element analysis, [ 18–20 ] and k ‐means [ 21–23 ] are the most commonly used data‐driven monitoring methods. Patan and Korbicz implemented recurrent neural networks as a one‐step‐ahead predictor and proposed a nonlinear model to analyze the performance of predictive control methods in boiler units.…”

Section: Introductionmentioning

confidence: 99%

Layered monitoring of xylenol tail gas treatment process based on stationary subspace analysis

Luan

Liu

2022

Can J Chem Eng

View full text Add to dashboard Cite

To deal with the inaccurate monitoring caused by the non-stationary characteristics of the xylenol tail gas treatment process, a layered monitoring scenario utilizing stationary subspace analysis has been presented in the present study. First, principal component analysis (PCA) is applied to establish the upperlevel stationary monitoring model. Then the sliding time window is used to capture the dynamic information of the remaining non-stationary features, update the model, and establish the lower non-stationary monitoring model.Next, the Bayesian information criterion is employed to build the global monitoring model according to the detection results of the upper and lower layers. Finally, the industrial data collected from the industrial boiler is used to evaluate the effectiveness and applicability of the proposed method. It is demonstrated that layered monitoring has higher accuracy than conventional non-stationary process monitoring techniques. Moreover, it can capture the faults in the stationary features masked by the non-stationary features.

show abstract

“…Lamiaa M. Elshenawy, Mohamed A, et al established a fault monitoring and diagnosis system for nuclear power plants using unsupervised machine learning [10]. Khalid, Salman, Lim, Woocheol, et al used the machine learning method to detect the leakage of boiler water wall tubes in steam power plants [11].…”

Section: Introductionmentioning

confidence: 99%

Discriminability Analysis of Characterization Parameters in Micro-Leakage of Turbocharged Boiler’s Evaporation Tube

Xia

Geng

et al. 2022

Energies

View full text Add to dashboard Cite

It is extremely dangerous for a turbocharged boiler to have a leakage fault in its vaporization tube. However, early detection and fault diagnosis of micro-leakage faults are very difficult. On the one hand, there are few fault samples that lead to a difficult and intelligent diagnosis. On the other hand, the system fault response characteristics of the characterization parameters in the process are complex and easily confused with the load-changing characteristics. In order to obtain fault samples and identify fault characteristics, a fault simulation model for the micro-leakage of the boiler evaporation tube is established based on the dynamic mathematical model of all working conditions. The model’s effectiveness is verified by typical fault experiments. The dynamic simulation experiments of three kinds of micro-leakage and four kinds of load changing were carried out. Through the analysis of combustion equilibrium and vapor-liquid equilibrium of 14 groups of characterization parameters, it is found that: (1) The reason for the poor discriminability in micro-leakage faults is that most of the characterization parameters tend to balance after 300 s and the dynamic response characteristics are similar to those of load increase. (2) There are four highly distinguishable parameters: the speed of the turbocharger unit, the air supply flow, the flue gas temperature at the superheater outlet, and the furnace pressure. When the micro-leakage fault is triggered, the first three parameters have a large disturbance. They show a trend of decreasing first and then increasing in short periods, unlike normal load-changing conditions. The fourth parameter (furnace pressure) rises abnormally fast after failure. (3) Under the normal working condition of varying loads, the main common parameters take 300 s to stabilize; the common stability parameter values should be recorded because when the micro-leakage fault of evaporation occurs, the steady-state increment of failure is larger than the normal steady increment under variable load conditions, by 2 to 3 times. (4) As the leakage fault increases, the disturbance amplitude of the characteristic parameters becomes larger. In addition, the stability of the steam system becomes worse, and fault discrimination becomes more obvious.

show abstract

Intelligent Steam Power Plant Boiler Waterwall Tube Leakage Detection via Machine Learning-Based Optimal Sensor Selection

Cited by 17 publications

References 48 publications

A Machine Vision—Based Pipe Leakage Detection System for Automated Power Plant Maintenance

A Machine Vision—Based Pipe Leakage Detection System for Automated Power Plant Maintenance

Layered monitoring of xylenol tail gas treatment process based on stationary subspace analysis

Discriminability Analysis of Characterization Parameters in Micro-Leakage of Turbocharged Boiler’s Evaporation Tube

Contact Info

Product

Resources

About