On-line measurement of temperature and CO2 concentration in flue gas based on deep learning coupled emission spectroscopy
Based on low-resolution infrared spectroscopy acquisitionand deep learning computational methods,an online detection method for flue gas temperature and CO2 concentration is proposed.The gas spectral radiation model was used to calculate the training data,the distribution of flame flue gas temperature and CO2 concentration was inverted based on a multi-layer perceptron(MLP)neural network.Results show that the inversion errors of the MLP neural network model for temperature and CO2 and H2O volume fractions are less than 1％,and the prediction accuracies are all greater than 94.5％,which has good generalization and prediction capabilities.A set of on-line detection device for flue gas temperature and CO2 concentration based on deep learning coupled with emission spectroscopy was estab-lished,and the ethylene diffusion flame and C2H4/NH3 partially premixed flame were investigated.The measurement results of flue gas temperature and CO2 volume fraction for the ethylene diffusion flame were consistent with the simulated flame results,which verified the feasibility of the online detection method based on deep learning coupled with emission spectroscopy.Changing the ammonia doping ratio of the partially premixed flame and analyzing the temperature and CO2 concentration changes of the gas at different heights above the central axis of the flame,results show that the flue gas temperature at the same height increases with the increase of the doped ammonia,and the CO2 volume fraction shows a tendency to increase and then decrease sharply.The proposed method can detect the changes of temperature and CO2 more sensitively,which can be used for combustion diagnostic studies of many kinds of flames,and also has some potential appli-cations in the online detection of carbon emissions in power plants.
flameCO2on-line measurementinfrared spectroscopy analysismulti-layer perceptrongas temperature
NETL
NSTL
万方数据
