| Automation of the prediction of the end temperature of BOF is a difficult problem in the process of steel making. At present, only through the furnace gas analysis method or artificial experience to predict. But the furnace gas analysis method is expensive and the maintenance cost is high, and the manual experience is low. Due to the non-contact measurement of the optical radiation temperature measurement, it is easy to maintain. But due to the non blackbody flame properties, blackbody radiation formula can not be used directly. Based on the Planck formula and the Wayne approximation, dual wavelength temperature measurement formula is introduced and the measurement formula is deduced. In view of the above problems, based on the formula of the color temperature, combined with spectral data got by fiber optic spectrometer, loss form assumptions on the spectral intensity is made to build the model. Analogy experiment is used to derive the expression form of the sum term. Combined with a spectral data to solve the temperature, a new model for judging the final temperature is established. In the application of the model, considering the complex production environment, there is noise in spectral data and the temperature data. So the data were processed by FFT filtering and continuous smoothing filtering method. The results were more accurate, and the the spectral temperature measurement model was corrected. Set up the hardware platform using the telescope system, fiber optic spectrometer, optical fiber and computer. Using C, MATLAB and Lab View to write the software program in the field and simulation environment. The modified spectral temperature measurement model is applied to the system. The temperature measurement system is tested in the field, and the hit rate of 20℃ is above 90%. The results show that this system fully meets the requirements of practical production, and has important significance to the improvement of production efficiency. |
PDF Full Text Request