Research on Flame Radical Imaging and Extreme Learning Machine to Prediction of NOx Emissions

Abstract

Abstract: Flame radicals are crucial for an in-depth understanding of the combustion mechanisms. The spectral characteristics of flame radicals were studied based on digital imaging and feature extraction techniques. The information obtained was used to establish the extreme learning machine (ELM) model which can be used to predict the NOx emissions based on the experimental data and digital simulation from a biomass-gas-air combustion process. The digital images of four flame radicals, i.e., OH*, CN*, CH* and C₂*, were collected using an EMCCD (Electron Multiplying Charge Coupled Device) camera. The image segmentation was performed using the fuzzy C-means (FCM) algorithm, and image features were extracted. Finally, the ELM model was built for the prediction of NOx emissions based on the radical features and flame temperture. The experimental data on a gas-biomass combustion test rig demonstrate the validity of the proposed ELM model.

Key words: flame radical, digital image processing, extreme learning machine, NOx emissions, prediction

CLC Number:

TP183

Li Xinli, Li Nan, Sun Yujia, Lu Gang, Yan Yong, Liu Shi. Research on Flame Radical Imaging and Extreme Learning Machine to Prediction of NOx Emissions[J]. Journal of System Simulation, 2016, 28(5): 1179-1185.

References

[1] 田贺忠, 赵丹, 王艳. 中国生物质燃烧大气污染物排放清单[J]. 环境科学学报, 2011, 31(2): 349-357.
[2] 冯磊华, 桂卫华, 杨峰. 基于改进PSO的电站锅炉低NOx燃烧优化[J]. 系统仿真学报, 2011, 23(12): 2812-2815.
(Feng L H, Gui W H, Yang F.Improved Particle Swarm Optimization to Optimize Combustion of Power Plant Boiler for Low NOx Emission[J]. Journal of System Simulation (S1004-731X), 2011, 23(12): 2812-2815.)
[3] Zhou H, Cen K F, Fan J R.Multi-objective optimization of the coal combustion performance with artificial neural networks and generic algorithms[J]. International Journal of Energy Research (S0363-907X), 2005, 29(6): 499-510.
[4] Zhou H, Cen K F, Fan J R.Modelling and optimization of the NOx emission characteristics of a tangentially fired boiler with artificial neural networks[J]. Energy (S0360-5442), 2004, 29: 167-183.
[5] 王淅芬, 罗自学, 周怀春. 基于炉内温度分布的NOx排放特性的神经网络模型[J]. 能源技术, 2009, 30(3): 133-136.
[6] 王春林, 周昊, 李国能, 等. 基于遗传算法和支持向量机的低NOx燃烧优化[J]. 中国电机工程学报, 2007, 27(11): 40-44.
[7] 吕游, 刘吉臻, 杨婷婷, 等. 基于PLS特征提取和LS-SVM结合的NOx排放特性建模[J]. 仪器仪表学报, 2013, 34(11): 2418-2424.
[8] 秦翠翠, 贾立. 集成软测量方法研究及其在烟气含氧量中的应用[J]. 系统仿真学报, 2014, 26(7): 1497-1502.
(Qin C C, Jia L.Research of Method of Integrating Soft Measurement for Flue Gas Oxygen Content[J]. Journal of System Simulation (S1004-731X), 2014, 26(7): 1497-1502.)
[9] Li K, Thompson S, Wieringa P A, et al.Neural networks and genetic algorithms can support human supervisory control to reduce fossil fuel power plant emissions[J]. Cognition, Technology & Work (S1435-5558), 2003, 5: 107-126.
[10] Ahamd A L, Azid I A, A. Yousof R, et al. Emission control in palm oil mills using artificial neural network and genetic algorithm[J]. Computers & Chemical Engineering (S0098-1354), 2004, 28: 2709-2715.
[11] Shakil M, Elshafei M, Habib M A, et al.Soft sensor for NOx and O₂ using dynamic neural networks[J]. Computers & Electrical Engineering (S0045-7905), 2009, 35: 578-586.
[12] 翟晓东, 丁艳军, 彭志敏, 等. 利用OH自由基发射光谱测量氧炔焰燃烧温度[J]. 清华大学学报(自然科学版), 2012, 52(7): 980-983.
[13] Krabicka J, Lu G, Yan Y.Profiling and characterization of flame radicals by combining spectroscopic imaging and neural network techniques[J]. IEEE Trans. Instrumentation and Measurement (S0018-9456), 2011, 60(5): 1854-860.
[14] Higgins B, Mcquay M Q, Lacas F, et al.Systematic measurements of OH chemiluminescence for fuel-lean, high-pressure, premixed, laminar flames[J]. Fuel (S0016-2361), 2001, 80(1): 67-74.
[15] Higgins B, Mcquay M Q, Lacas F, et al.An experimental study on the effect of pressure and strain rate on CH chemiluminescence of premixed, fuel-lean methane/air flames[J]. Fuel (S0016-2361), 2001, 80(11): 1583-1591.
[16] Docquier N, Lacas F, Candel S.Closed-loop equivalence ratio control of premixed combustors using spectrally resolved chemiluminescence measurements[J]. Proc. Combust. Inst (S1540-7489), 2002, 29: 139-145.
[17] Hardalupas Y, Orain M, Panoutsos C S, et al.Chemiluminescence sensor for local equivalence ratio of reacting mixtures of fuel and air (MAST B LIQUD)[J]. Applied Therm. Engineering (S1359-4311), 2004, 24(11/12): 1619-1632.
[18] Bombach R, Kappeli B.Simultaneous visualization of transient species in flames by planar-laser induced fluorescence using a single laser system[J]. Applied Physics B:Lasers and Optics (S0946-2171), 1999, 68(2): 251-255.
[19] Li X L, Sun D, Lu G, et al.Prediction of NOx Emissions based on Flame Radical Imaging and Neural Network Techniques[C]// 2012 IEEE Imaging Science and Techniques (IST2012), Manchester, England. USA: IEEE, 2012, 6.
[20] Li N, Lu G, Li X L, et al.Prediction of pollutant emissions of biomass flames using digital imaging, contourlet transform and Radial Basis Function network techniques[C]// 2014 IEEE Instrumentation and Measurement Technology conference (I2MTC), Montevideo, Uruguay. USA: IEEE, 2014, 5.
[21] Giorgio Z.Flame Emission Spectroscopy: Fundamentals and Applications[M]// Lecture given at ICS Training course on laser diagnostics on combustion processes. NILES, Egypt: University of Cairo, 2000: 18-22.
[22] 杨铭. CN自由基势能曲线的研究 [D]. 上海: 华东师范大学, 2005, 1-4.
[23] 李旭超, 刘海宽, 王飞, 等. 图像分割中的模糊聚类方法[J]. 中国图象图形学报, 2012, 17(4): 447-458.
[24] Huang G B, Zhu Q Y, Siew C K.Extreme learning machine: Theory and applications[J]. Neurocomputing (S0925-2312), 2006, 70(1-3): 489-501.