Fig. 1. Schematic diagrams of radiative transfer paths for two types of one-dimensional cases for parameter optimization of MSMGWB model

Fig. 2. Schematic diagram of cumulative k-distribution function

Fig. 3. Influences of grouping strategy, Gaussian quadrature format, and reference temperature on MSMGWB model

Fig. 4. Genetic algorithm genotype and crossover process diagram

Fig. 5. Genetic algorithm iteration process diagram

Fig. 6. Offspring generation diagram of genetic algorithm

Fig. 7. Genetic algorithm iteration results. (a) Calculated results of 10 Pareto fronts; (b) iteration convergence history

Fig. 8. Comparison of results of genetic algorithm and exhaustive method

Fig. 9. ferr0 results of 100×400 grouping strategy combinations

Fig. 10. Process schematics of four iteration scanning schemes

Fig. 11. Performance comparison chart of four iteration scanning schemes

Fig. 12. Ratios between ferr0 results of different population sizes and a population size of 40000

Fig. 13. Correlation analysis for sample with population sizes of 5000 and 40000

Fig. 14. Optimization results of genetic algorithm in 3‒5 μmband. (a) Pareto front results; (b) comparison of performance between optimized MSMGWB model and MSMGWB model without optimization

Fig. 15. Genetic algorithm optimization results in 8‒14 μm band. (a) Pareto front results; (b) comparison of performance between optimized MSMGWB model and MSMGWB model without optimization

Fig. 16. Error distribution diagram of optimized MSMGWB model compared to LBL calculation for aircraft engine exhaust systems and jet remote infrared imaging scenes