Rescarch On Combustion Stability Based On Deep Learning And Flame Free Radicals

In the industrial production process,the combustion state of the boiler plays a vital role in the safety and stability of operation.In this paper,combustion stability is studied based on flame free radicals and in-depth learning.Four kinds of flame free radicals(CN*,CH*,C2*and OH*)images were acquired based on EMCCD enhanced camera.The concept of transfer learning was introduced.The feature extraction was carried out by depth learning algorithm.The combustion condition identification migration model was established based on the index interval index combustion stability decision-making method,and the combustion stability discrimination was realized.The main contents of this paper are as follows:Firstly,the principle of deep learning algorithm is introduced,and the advantages and disadvantages of different learning frameworks,including Alex Net.VGG Net and Inception Net,are compared and analyzed.Secondly,based on the laboratory combustion test device,different experimental conditions were designed to obtain the spectral information of four kinds of flame free radicals under different combustion conditions.According to the spectral information,the relationship between flame free radicals and combustion conditions was analyzed.Finally,the combustion stability decision-making method based on the interval index is used to construct the instability,and the combustion stability is divided into four stages to establish the criterion of combustion stability,and the migration learning is introduced to improve the inception-v3 model.Four kinds of flame free radical images under different combustion conditions were obtained by EMCCD enhanced camera.The image characteristic parameters were extracted by improved inception-v3-V3 model.The combustion conditions were identified by using Softmax classifier,and the combustion stability was judged.The simulation results show that the combustion stability discriminant model is effective,and the recognition rates of different free radicals are compared and analyzed.