一种基于加权多准则的火炮爆炸场中高温火球三维温度场的迭代重建算法

针对爆炸场温度高、温度动态范围大、不可直接测量、温度场重建难以取得较高精度等问题，改进了一种在图像重建中广泛应用的联合代数重建（simultaneous algebraic reconstruction technique，SART）算法，以实现爆炸场中温度场的重建。该改进方法将原算法迭代固定步长改为自适应步长，通过判断相邻像素间是否产生畸变来决定迭代步长的大小，无畸变时迭代步长较大，有畸变时则加入惩罚使迭代步长变小；并在SART算法满足的最小二乘准则中加入均匀性准则，提高重建质量。通过仿真实验讨论了惩罚区域和惩罚系数对重建图像质量的影响，并对比了传统SART算法和在校正项中加入惩罚的SART算法。实验结果表明，由于改进方法在迭代系数中加入了惩罚项，得到的重建图像质量更高。

An iterative reconstruction algorithm for the 3D temperature field of the high-temperature fireball in a gun explosion field based on weighted multi-criteria

Because of the high temperature, large temperature dynamic range and inability to make direct measurement, it is difficult to obtain high accuracy in the temperature reconstruction of an explosion field. This paper reports an improved simultaneous algebraic reconstruction technique (SART) algorithm with a wide range of application in image reconstruction to realize the reconstruction of the temperature field. This method replaces the fixed iteration step in the original SART with an adaptive step. The size of the iteration step is determined by judging whether or not there is a distortion between adjacent pixels. The iteration step is larger when there is no distortion, but otherwise the iteration step is reduced by adding a penalty if there is a distortion. To improve the quality of the reconstructed image, the uniformity criterion is combined with the least squares criterion in our improved SART algorithm. In simulation experiments the influence of the penalty area and the penalty coefficient on the quality of the reconstructed image is discussed. The improved method is compared with the traditional SART algorithm and the SART algorithm with a penalty added to the correction term. The experimental results show that the improved method reported in this paper can afford a better reconstructed image by means of adding a penalty term to the iteration coefficient.