Quantitative assessment method of muzzle flash and smoke at high noise level on field environment

It is quite a challenge to obtain the temperature and species concentration fields of muzzle flash at high noise level. In this numerical study, radiation intensity of muzzle flash received by the high-speed Complementary Metal-Oxide-Semiconductor (CMOS) camera was simulated based on the line-of-sight method in the direct radiative transfer problem. The inverse radiative transfer problem of reconstructing distributions of temperature and soot volume fraction from the knowledge of flame radiation intensity was transformed into a minimization optimization problem and a meta-heuristic algorithm was used to solve the problem. The effects of the number of detection lines, optical thickness and measurement errors on the reconstruction results were discussed in details. A method to estimate the noise level of radiation intensity was developed, experimental results showed that the signal-to-noise ratio (SNR) of radiation intensity can be successfully inferred when the SNR is greater than 20 dB. Subsequently, prior knowledge of the noise level was introduced in the regularization to achieve a meaningful approximation of the exact value. The reconstruction of the soot volume fraction filed with SNR greater than 40 dB is considered successful with the inclusion of an appropriate regularization term in the objective function, and the reconstruction of the temperature field is feasible even with SNR as low as 15 dB. The high tolerance to the noise level of the radiation intensity gives the reconstruction algorithm the potential to be used in practical experiments of muzzle flash.