Kömür Yakıcılarında Yanma Verimi Tahmini için Gömülü Platformda Çalışabilen Evrişimsel Sinir Ağının Parametre Analizi

Öz

Kömür yakıcılarında yanma veriminin doğru ve etkin bir şekilde hesaplanması endüstriyel kazan üreticileri için oldukça önemlidir. Kazan veriminin hesaplanabilmesi için iki temel yaklaşımın olduğu görülmektedir: 1) bacadan çıkan gazın analizi; 2) kazandaki yanma odasının görüntülenmesi. Kullanımı yeterince kolay olmayan baca gazı analizörleri yüksek maliyete sahiptir. Ayrıca baca ile yanma odası arasındaki fiziksel uzaklık yapılan ölçümün zaman gecikmeli olmasına neden olmaktadır. Yanma odasının görüntülenmesine dayalı yöntemler bahsedilen dezavantajları içermemektedir. Bu çalışmada önerilen ve yanma odasının görüntülenmesine dayanan sistemin literatüre iki katkısı bulunmaktadır: 1) yanma veriminin tahmininde ilk defa modern evrişimsel sinir ağları (ESA) yaklaşımının kullanılması; 2) Optimum parametrelere sahip ESA mimarisinin gömülü bir platformda çalışabilmesi. Klasik sınıflandırma teknikleri ve CPU destekli bir işlemci kartı kullanıldığında, 1,7 saniyede 1 adet alev formu görüntüsünden verim hesaplanabilirken, önerilen ESA mimarisi ve GPU destekli bir işlemci kartı kullanıldığında bu sayı saniyede yaklaşık 20 adet seviyesine çıkmaktadır (34 kat hızlı). Elde edilen sonuçlar, kömür kazanı yanma verimi tahmininde önerilen ESA mimarisinin ve donanımının klasik yaklaşımlara olan üstünlüğünü açık bir şekilde ortaya koymaktadır.

Anahtar Kelimeler

• Kömür
• Yakıcı Sistem
• Yanma Verimi
• Görüntü İşleme
• Evrişimsel Sinir Ağları

Parameter Analysis of Convolutional Neural Network Operated on Embedded Platform for Estimation of Combustion Efficiency in Coal Burners

Abstract

Accurately and effectively calculating combustion efficiency in coal burners is crucial for industrial boiler manufacturers. Two main approaches can be used to calculate boiler efficiency: 1) Analyzing the gas emitted from the flue; 2) Visualizing the combustion chamber in the boiler. Flue gas analyzers, which are not user-friendly, come with high costs. Additionally, the physical distance between the flue and the combustion chamber causes the measurement to be delayed. Methods based on visualizing the combustion chamber do not have these disadvantages. This study proposes a system based on visualizing the combustion chamber and has two contributions to the literature: 1) for the first time, the modern Convolutional Neural Networks (CNN) approach is used to estimate combustion efficiency; 2) the CNN architecture with optimal parameters can work on an embedded platform. When classical classification techniques and a CPU-supported processor card are used, efficiency can be calculated from one flame image in 1.7 seconds, while this number increases to approximately 20 frames per second (34 times faster) when the proposed CNN architecture and GPU-supported processor card are used. The results obtained demonstrate the superiority of the proposed CNN architecture and hardware over classical approaches in estimating coal boiler combustion efficiency.

Keywords

• Coal
• Combustor
• Combustion Efficiency
• Image Processing
• Convolutional Neural Networks

Kaynakça

1. Onat C, Talu MF, Daskin M, Mercimek M. Otomatik Beslemeli Kömür Kazanlarinda Alev Formu İle Yanma Verimi Arasindaki İlişkinin İncelenmesi. Mühendis ve Makine 2015; 669: 70–79, 2015.
2. Testo SE, KGaA Co. Baca gazı ölçüm cihazları. 2019 [cited 2019 October 30. Available from: https://www.testo.com/tr-TR/ueruenler/gaz-oelcuem-cihazlatri.
3. Lee CL, Jou CJG. Saving fuel consumption and reducing pollution emissions for industrial furnace. Fuel Process. Technol. 2011; 92(12):2335–2340.
4. Hao Z, Kefa C, Jianbo M. Combining neural network and genetic algorithms to optimize low NOx pulverized coal combustion. Fuel: 2011; 80 (15): 2163–2169.
5. Hao Z, Qian X, Cen K, Jianren F. Optimizing pulverized coal combustion performance based on ANN and GA. Fuel Process. Technol. 2004; 85 (2–3):113–124.
6. Zheng Z, Yao M. Charge stratification to control HCCI: Experiments and CFD modeling with n-heptane as fuel. Fuel. 2009; 88(2): 354–365.
7. Liu D, Yan J, Wang F, Huang Q, Chi Y, Cen K. Experimental reconstructions of flame temperature distributions in laboratory-scale and large-scale pulverized-coal fired furnaces by inverse radiation analysis. Fuel. 2012; 93: 397–403.
8. Kuprianov V, Chullabodhi C, Jornjumrus W. Cost based optimization of excess air for fuel oil/gas-fired steam boilers. RERIC Int. Energy J.1999; 21(2): 83–91.

9. Yamaguchi T, Grattan KTV, Uchiyama H, Yamada T. A practical fiber optic air-ratio sensor operating by flame color detection. Rev. Sci. Instrum. 1997; 68(1): 197–202.
10. Lino N, Tsuchino F, Yano T. TIMEWISE VARIATION OF TURBULENT JET DIFFUSION FLAME SHAPE BY MEANS OF IMAGE PROCESSING. J. Flow Vis. Image Process.1998; 5(4): 275–281.
11. Huang Y, Yan Y, Lu G, Reed A. On-line flicker measurement of gaseous flames by image processing and spectral analysis. Meas. Sci. Technol. 1999; 10(8): 726–733.
12. Golgiyaz S, Talu MF, Daşkın M, Onat C. Estimation of excess air coefficient on coal combustion processes via gauss model and artificial neural network. Alexandria Eng. J.2022; 61(2): 1079–1089. doi: 10.1016/j.aej.2021.06.022.
13. Baek WB, Lee SJ, Baeg SY, Cho CH. Flame image processing & analysis for optimal coal firing of thermal power plant. ISIE 2001 IEEE Int. Symp. Ind. Electron. proceedeing, Vols I-III. 2001; 928-931.
14. Eleyan A, Dem H. Co-occurrence matrix and its statistical features as a new approach for face recognition. Comp Sci. 2011; 19(1).
15. TALU MF, ONAT C, DASKIN M. Prediction of Excess Air Factor in Automatic Feed Coal Burners by Processing of Flame Images. Chinese J. Mech. Eng. 2017; 30(3): 722–731.
16. Chen LC, Papandreou G, Schroff F, Adam H. Rethinking Atrous Convolution for Semantic Image Segmentation. 2017. arXiv:1706.05587
17. Chen J, Chan LLT, Cheng YC. Gaussian process regression based optimal design of combustion systems using flame images. Appl. Energy. 2013; 111: 153–160.
18. Huang B, Luo Z, Zhou H. Optimization of combustion based on introducing radiant energy signal in pulverized coal-fired boiler. Fuel Process. Technol. 2010; 91(6): 660–668. doi: 10.1016/j.fuproc.2010.01.015.
19. Huang B, Luo Z, Zhou H. Optimization of combustion based on introducing radiant energy signal in pulverized coal-fired boiler. Fuel Process. Technol. 2010; 91(6): 660–668.
20. Krizhevsky A, Sutskever I, Hinton GE. ImageNet Classification with Deep Convolutional Neural Networks. 2012; 25(2). doi:10.1145/3065386.
21. Girshick R, Donahue J, Darrell T, Malik J. Rich feature hierarchies for accurate object detection and semantic segmentation. Nov. 2013.
22. Farabet C, Couprie C, Najman L, LeCun Y. Learning Hierarchical Features for Scene Labeling. IEEE Trans. Pattern Anal. Mach. Intell. 2013; 35 (8): 1915–1929.
23. ÇALIŞAN M, TALU MF. Comparison of Methods for Determining Activity from Physical Movements. J. Polytech. 2020; 0900(1): 17–23.
24. Celik G. Detection of Covid-19 and other pneumonia cases from CT and X-ray chest images using deep learning based on feature reuse residual block and depthwise dilated convolutions neural network. Appl. Soft Comput.2023; 133; 109906. doi: 10.1016/j.asoc.2022.109906.
25. Çelik G, Talu MF. A new 3D MRI segmentation method based on Generative Adversarial Network and Atrous Convolution,. Biomed. Signal Process. Control. 2022; 71: 103155. doi: 10.1016/j.bspc.2021.103155.
26. Başaran E. A new brain tumor diagnostic model: Selection of textural feature extraction algorithms and convolution neural network features with optimization algorithms. Comput. Biol. Med. 2022; 148: 105857. doi: 10.1016/j.compbiomed.2022.105857.
27. Rawal V, Prajapati P, Darji A. Hardware implementation of 1D-CNN architecture for ECG arrhythmia classification. Biomed. Signal Process. Control. 2023; 85; 104865. doi: 10.1016/j.bspc.2023.104865.
28. Hamdi S, Oussalah M, Moussaoui A, Saidi M. Attention-based hybrid CNN-LSTM and spectral data augmentation for COVID-19 diagnosis from cough sound. J. Intell. Inf. Syst. 2022; 59(2): 367–389. doi: 10.1007/s10844-022-00707-7.
29. Polat K, Nour M. Parkinson disease classification using one against all based data sampling with the acoustic features from the speech signals. Med. Hypotheses. 2020; 140:109678. doi: 10.1016/j.mehy.2020.109678.
30. Golgiyaz S, Talu MF, Onat C. Artificial neural network regression model to predict flue gas temperature and emissions with the spectral norm of flame image. Fuel, 2019; 255; 115827. doi: 10.1016/j.fuel.2019.115827.