基于多光谱吸收的交替迭代优化算法

Alternating Iterative Optimization Algorithm Based on Multispectral Absorption

基于模拟退火算法的吸收光谱技术可以实现高温气体温度浓度的二维分布重建。但其计算效率过低,不适合于大规模的场反演计算。针对于该问题,提出了一种基于多光谱吸收的温度场和浓度场重建的交替迭代优化算法。首先对浓度场的计算提出了线性化方案,将原系统半线性化,从而弱化了原系统的非线性;其次对待测温度场、浓度场提出了交替迭代方案,每一步迭代降低一半的未知变量个数;最后用优化技术引入惩罚项来克服重建过程的病态性,在每一步迭代时能自动调整正则化参数,保证了迭代解收敛于精确解。仿真和实验结果表明该算法具有较强的抗噪声能力,在保持重建精度的前提下,能够显著的缩短计算时间,适用于大规模的场反演计算。

The two dimensional distribution of concentration and temperature can be achieved by simulated annealing algorithm for Spectral Absorption technology. However, its computation efficiency is too low, not, suitable for large scale inversion calculation. An alternately iterative reconstruction algorithm is proposed to solve this problem. Firstly, a linearization scheme was used in calculation of concentration field, weakening the nonlinearity of the original system; secondly, an alternately iterative scheme is used to reduce half the unknown variable number in each iteration;Finally, a penalty term is introduced in the optimization process to overcome the illness problem of reconstruction. The simulation and experiment results show that the algorithm is capable of good inversion accuracy, and it can significantly reduce the computation cost and immune to strong noise.Further-more, it is suitable for large-scale inversion calculation.