| In order to achieve the "double carbon goal" and establish a diversified energy structure,the proportion of clean energy in the energy structure is increasing.But its randomness and volatility greatly affect the safe and stable operation of the power grid,coal power is bound to assume the role of peak regulation.However,the operation of coal-fired boiler is not stable and even serious accidents may occur when the boiler is operated under peak regulating state.Boiler temperature can directly reflect the running state of the boiler,so the accurate measurement of boiler temperature has important reference significance for regulating and controlling the boiler.At present,many temperature detection technologies of power plant boilers have been developed,but because of the harsh industrial environment of power plant boilers,acoustic wave method,laser spectroscopy and other technologies are challenging.Based on this,radiation thermometry develops rapidly because of its advantages of low hysteresis and wide range of temperature measurement,especially in recent years with the development of technology and widely used.Based on flame radiation transfer,this thesis combines flame image processing technology with radiation transfer inverse problem to reconstruct the temperature field and radiation parameters in the furnace section.Specific work is as follows:In this thesis,the quantitative function relationship between flame temperature and flame image is obtained by solving the radiation transfer equation on the premise of ignoring flame scattering,self-absorption and wall emission and reflection,and the flame radiation imaging model is established.This chapter firstly studies the influence of different reconstruction algorithms on the reconstruction accuracy of temperature and soot concentration: Tikhonov regularization algorithm is better than QR decomposition LSQR algorithm in reconstruction accuracy and noise resistance.Secondly,the influence of cameras on the reconstruction accuracy of temperature and carbon smoke concentration is analyzed.Increasing the number of cameras can effectively improve the reconstruction accuracy.Compared with temperature reconstruction,soot concentration reconstruction requires more cameras for better reconstruction results.In order to reconstruct better temperature and soot concentration distribution at the same time,more than six cameras are required to collect flame images,and the regularization algorithm based on Tikhonov is used for inversion calculation.Thirdly,the influence of measurement noise on the accuracy of temperature and soot concentration reconstruction is studied.The influence of measurement noise on soot concentration is far greater than that on temperature.The regularization algorithm based on Tikhonov has good robustness.Even if the signal-to-noise ratio is 39 d B,the reconstruction error of the regularization algorithm based on Tikhonov is only about4%.Finally,the effect of camera interval Angle error on the reconstruction effect is discussed.The effect of camera interval Angle error is very great.In practical experiments,the camera interval Angle error should be minimized to improve the reconstruction effect.Secondly,a two-dimensional rectangular furnace system consisting of emitting and reflecting walls and emitting and scattering space media is studied.The radiation imaging model was calculated by DRESOR method,and the quantitative function relationship between flame image intensity,temperature and radiation parameters in furnace was established.Based on Tikhonov’s regularization algorithm and optimization algorithm,the temperature field and radiation parameters in the furnace are reconstructed simultaneously.Firstly,the influence of the assumption of radiation parameter uniformity on the temperature reconstruction in the furnace is discussed.Although the radiation parameters in the furnace are not uniformly distributed,it is still feasible to reconstruct the temperature distribution in the furnace by using the radiation parameter uniformity.The maximum relative error of temperature reconstruction is5.5%,which meets the demand of industrial temperature measurement.Secondly,the influence of the number of cameras on temperature reconstruction is discussed.Increasing the number of cameras can effectively improve the effect of temperature reconstruction.When eight cameras are used,the effect of temperature reconstruction has reached the demand of industrial temperature measurement,and the maximum relative error of temperature reconstruction is 2.2773%.Finally,the influence of measurement noise on temperature reconstruction is analyzed.The algorithm has good robustness,even if the measurement noise is 20,the maximum relative error of temperature reconstruction is still about 10%.Finally,an experimental study on the temperature and radiation parameters of a pulverized coal boiler was carried out.During the research,a new calibration method is proposed,and the variable function relationship between radiation intensity,image intensity and camera exposure time is established.This method can eliminate the influence of the change of exposure time on the calibration result and improve the calibration accuracy.The maximum relative error of temperature reconstruction is 4%.During the experiment,the flame image detector was first used to collect the flame image through the fire viewing hole,and the temperature distribution and radiation parameters of the section of the pulverized coal boiler under different loads were reconstructed.The experimental results showed that the boiler load decreased from159 MW to 150 MW,and the maximum temperature reconstruction also decreased from1785 K to 1535 K to 1516 K.The reconstructed mean temperature changes from 1672 K to 1421 K and finally reaches the minimum value of 1417 K.With the increase of furnace temperature,the equivalent absorption coefficient and scattering coefficient of furnace increase. |
PDF Full Text Request