Research On The Reconstruction Method Of The Computed Tomography Of Chemiluminescence

The combustion flow inside the engine is extremely complex,and the mechanism is still unclear.The existing combustion diagnosis technology cannot realize the instantaneous three-dimensional measurement of turbulent combustion process.The chemiluminescence of a flame can reflect combustion state and flame structure information.Computed tomography(CT)technology can reconstruct the spacial structure of objects.Combining the chemiluminescence properties of flame with CT technology to form the computed tomography of chemiluminescence(CTC)technique,which can realize instantaneous three-dimensional measurement of the combustion flow field.In this dissertation,the reconstruction method of CTC is mainly studied,and the radiation projection model and iterative reconstruction algorithm are improved by combining with the characteristics of flame.The effectiveness of the improved method is verified through a series of simulation and experimental studies.First,the light projection model in CTC technology is studied.Based on the principle of lens imaging and Monte Carlo particle tracking,a radial projection model is constructed and compared with the parallel light projection model.The process of solving the weight coefficient is optimized,and an improved weight coefficient calculation model is proposed.Research shows that the radiant projection model is more in line with the actual process and the improved weight coefficient model can obtain accurate reconstruction results.Then,the reconstruction algorithm is studied.The virtual flame model is constructed and the evaluation index is proposed.Meanwhile,an improved IART algorithm is proposed,and the reconstruction performance of several algorithms is compared.The various factors affecting the reconstruction accuracy are analyzed and simulated.The research finds that the improved algorithm not only can obtain high-precision reconstruction results,but also greatly improve the reconstruction speed.The more number of the projection angles,or the more uniform distribution,the higher the reconstruction accuracy.The larger the relaxation factor is,the smaller the reconstruction error will be.But if it is too large,the iteration will diverge.The projection resolution is too large or too small will affect the reconstruction accuracy.The improved algorithm has a certain anti-noise ability,and the reconstruction error increases linearly with the increase of noise intensity.Finally,a pilot verification study of CTC technology was carried out.A computing platform is built based on Matlab,and the experimental reconstruction of CTC technology is realized through projection preprocessing,camera position calibration,weight coefficient solving,iterative reconstruction and post-processing of results.Thefeasibility of CTC technology is verified by comparing the reconstructed image of the two-dimensional premixed flame with the actual captured image.The three-dimensional chemiluminescence field was reconstructed to analyze the flame structure.The research shows that the improved method proposed in this dissertation is feasible and can be used to reconstruct the real flame accurately.