Toward Database Development And Bayesian Optimization Algorithm Application Of Frequency Selective Surface Microwave Absorbers

The Radar Cross-Section(RCS)reduction using absorbing materials is one of the important measures to achieve radar stealth.Frequency selective surface(FSS),as a new type of structural absorbing material,plays an important role in RCS reduction.There are many structural parameters affecting the electromagnetic performance of FSS.How to effectively analyze the parameters of FSS absorbing structure and design the FSS structure to meet the specific target frequency band has become a hot topic of scholars.In this thesis,a database for frequency selective surface absorbers is developed,and the related data analysis functions are extended.The Bayesian optimization algorithm is studied and the optimization of single-ring frequency selective surface absorber is completed,which provides technical foundation for the design of frequency selective surface absorbers..Based on the database,the passive frequency selective surface microwave absorbers and the active frequency selective surface microwave absorbers data processing module are completed.The data processing of the passive frequency selective surface is mainly through performance evaluation of the passive model data in the database.The processing of the active FSS is to superimpose a plurality of frequency response states to obtain a reflectivity envelope.Through the dimension reduction technique,the influences of the variation of the absorbing structure variables on the absorbing properties,the bandwidth,the amplitude of the absorption peak,and the frequency of the absorption peak are analyzed,and the design law of the FSS absorber is obtained.The mathematical principle of Bayesian optimization algorithm(BOA)is studied.It creates a function distribution model by means of probability statistics,finds the position of the optimal solution in the model,which greatly improves the optimization efficiency.In order to compare the efficiency of the algorithm and the genetic algorithm,a single-ring frequency selective surface absorbing structure is designed,and the target frequency band is optimized to 4-8 GHz.The results show that both algorithms achieve the optimization goal,but the optimization efficiency of BOA is higher.Bayesian optimization algorithm has application significance in frequency selective surface optimization.