Support vector machine based online coal identification through advanced flame monitoring

This paper presents a new on-line coal identification system based on support vector machine (SVM) to achieve on-line coal identification under variable combustion conditions. Four different coals were burnt in a 0.3 MW coal combustion furnace with different coal feed rates, total air flow rates and flow rate ratios of primary air and secondary air. The flame monitoring system was installed at the exit of the burner to acquire the coal flame images and light intensity signals. Spatial and temporal flame features were extracted for coal identification. The averaged prediction accuracy is 99.1%. The mean value of the infrared signal has the most significant influence on prediction accuracy. For “unstudied” operation cases, the prediction accuracy is 94.7%.

[1]  Yong Yan,et al.  On-line fuel identification using digital signal processing and fuzzy inference techniques , 2004, IEEE Transactions on Instrumentation and Measurement.

[2]  Huaichun Zhou,et al.  Optimization of combustion based on introducing radiant energy signal in pulverized coal-fired boiler , 2010 .

[3]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[4]  Yong Yan,et al.  A digital imaging based multifunctional flame monitoring system , 2003, IEEE Transactions on Instrumentation and Measurement.

[5]  Yong Yan,et al.  Characterisation of biomass and coal co-firing on a 3 MWth Combustion Test Facility using flame imaging and gas/ash sampling techniques , 2009 .

[6]  Yong Yan,et al.  Online fuel tracking by combining principal component analysis and neural network techniques , 2005, IEEE Transactions on Instrumentation and Measurement.

[7]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[8]  Hua Wei Huang,et al.  Flame colour characterization in the visible and infrared spectrum using a digital camera and image processing , 2008 .

[9]  Fei Wang,et al.  The research on the estimation for the NOx emissive concentration of the pulverized coal boiler by the flame image processing technique , 2002 .

[10]  Yong Yan,et al.  Characterisation of an oxy-coal flame through digital imaging , 2010 .

[11]  Hao Zhou,et al.  Modeling and optimization of the NOx emission characteristics of a tangentially fired boiler with artificial neural networks , 2004 .

[12]  Yong Yan,et al.  On-line flicker measurement of gaseous flames by image processing and spectral analysis , 1999 .

[13]  Yong Yan,et al.  Independent Component Analysis–Based Fuel Type Identification for Coal-Fired Power Plants , 2012 .

[14]  Hao Zhou,et al.  Flame stability monitoring and characterization through digital imaging and spectral analysis , 2011 .

[15]  Yong Yan,et al.  Vision-based measurement of temperature distribution in a 500-kW model furnace using the two-colour method , 2000 .

[16]  A. Gil,et al.  Coal flame characterization by means of digital image processing in a semi-industrial scale PF swirl burner , 2012 .

[17]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[18]  Huaichun Zhou,et al.  A simple measurement method of temperature and emissivity of coal-fired flames from visible radiation image and its application in a CFB boiler furnace , 2009 .