Simultaneous Measurement Of Three-dimensional Temperature And Soot Concentration Distributions Of Flame Based On The Multi-light Field Camera System

Combustion is an extremely complex process involving the coupling and interaction of physical and chemical changes.The flame temperature and soot concentration distributions are the most important parameters to determine the state of matter and the process of the combustion reaction.The precise measurement of their distribution is not only conducive to revealing the essence of the combustion phenomenon and the laws of the combustion process,but also of great significance for improving energy conversion efficiency,controlling pollutants,and reducing carbon emissions.In view of the shortcomings of the existing simultaneous measurement systems of flame temperature and soot concentration distributions are complex and complicated to calibrate and adjust,combined with the advantages of light field imaging technology,such as compact and non-invasive system structure,this dissertation proposes a simultaneous measurement method for the flame temperature and soot concentration distributions based on the multi-LF imaging technique.Systematic theoretical,simulation and experimental studies are carried out.Firstly,the characteristics of different light field cameras(LFCs)are systematically analyzed.Based on these characteristics,the flame radiation imaging model of the multi-LFC system is proposed.To address the low-rankness and directional redundancy in flame radiation LF sampling,a novel approach i.e.,feature rays under-sampling(FRUS)is proposed to optimize the LF sampling.The performance of the FRUS approach is evaluated from the following aspects,namely,under-sampling methods,anti-noise ability,LFC parameters,and flame dividing voxels.The results indicate that the FRUS approach does not depend on the physical parameters of the flame.FRUS approach greatly improves the reconstruction efficiency and reduces the reconstruction error.For the single LFC and multi-LFC systems,the reconstruction time for the optimized sampling is about 1/20～1/10 and 1/15～1/10 of the original sampling.To improve the reconstruction quality of the flame temperature field in a confined optical space,a method is proposed by integrating a Quantum-behaved Particle Swarm Optimization algorithm to optimize the angular arrangement of the multi-LFC system.The stability and applicability of the proposed method are evaluated by numerical simulations.The effect of angular arrange optimization on the flame temperature reconstruction is quantitatively analyzed from the following aspects,namely,angular arrangement methods,the number of cameras,and flame structures.The results show that compared with other angular arrangements,the ill-conditioning of the constructed radiation transfer equation with the optimum angular arrangement is the weakest and the reconstruction quality is the best,which provides a theoretical basis for the multi-LFC system configuration in subsequent experimental studies.To address the problem of large reconstruction error caused by wide dynamic ranges of flame radiation signals,a method based on Weighted Non-Negative Least Squares(WNNLS)is introduced to recover the flame temperature and soot concentration distributions simultaneously.The proposed method improves the equality of LF flame radiation signals in the reconstruction process by assigning different weights to the flame radiation intensities,thereby reducing reconstruction errors.Both single-modal and bimodal flames are considered.The anti-noise ability and reliability of the Non-Negative Least Square(NNLS)and WNNLS algorithms are compared.The studies manifest that compared with the NNLS algorithm,the WNNLS algorithm reduces the number of iteration calculations and time,and has higher reconstruction accuracy.To verify the feasibility of the simultaneous measurement method of flame temperature and soot concentration distributions based on the multi-LFC system,experimental studies of the simultaneous measurement of ethylene diffusion flame temperature and soot concentration distributions are carried out.The integrated measurement system of multi-LFC is constructed.The multi-camera synchronous acquisition software is developed and the radiation intensity calibration of the photosensor is carried out to realize the synchronous acquisition of multi-angle flame radiation LF images and reliable acquisition of flame radiation information.The reconstruction experiments of flame temperature and soot concentration distributions are carried out on a multi-nozzle burner with different flame structures,and the reconstructed temperatures are compared with thermocouple measurement results.The results show that the reconstruction results of flame temperature are reliable and in good agreement with the thermocouple measurement results with a maximum difference of 77 K.The reconstructed concentration distributions accord with the evaluation law of soot.In conclusion,the proposed simultaneous measurement method for the flame temperature and soot concentration distributions based on the multi-LFC system is feasible and reliable.