A Study On The Optimization And Design Based On The Evolutionary Algorithms Of Absorbers

In this paper, the niching genetic algorithm with elitist model (NGA) and the team progressalgorithm (TPA) are used to optimize the multilayer coating and the textured absorbers. Theminimization of maximum reflection coefficient and average reflection coefficient as performanceindices is used as objective function for optimization.For the multilayer absorber design, the four-layer coating is only considered. The optimal desginresults of the multilayer with the two objective functions for all the two algorithms is analyzed.Comparative analysis of the simulated optimization results show that the optimal objective functionvalue is almost the same for the both optimization algorithms and prove that the results are credible.The results are also show that while the four-layer coating absorber constructed by using themaximum reflection coefficient as the objective function has a higher average reflection than usingaverage reflection coefficient as the objective, but it has a much lower worst reflection for both TEand TM polarizations.The design of the textured absorber with the maximum reflection coefficient as objectivefunction with both two evolutionary algorithms is analyzed. The method based on combined Matlabprogram and the simulation software HFSS has been presented in this part. The simulated results ofthe simualtion software provides the objective function value of the algorithms. From the simulatedoptimization results,the following conclusion becomes clear. Both TPA and NGA find almost thesame minimum reflection coefficient for both TE and TM polarization.This implies that the resultsare credible even these optimized design parameters are a little different. And the TPA returns theresult with fewer function evaluaions than NGA.TPA convergenes using about two-thirds the CPUtimes of NGA.