Robust Optimization Design Of Antenna Based On Adaptive Meta-Model

The antenna is a key electronic equipment for receiving and transmitting electromagnetic wave information.With the increasing utilization of antennas,the requirement for antennas increases sharply,which making antennas development to the direction of miniaturization,wideband,high-gain and high-density.The antenna is a high-precision device,and the existing processing technology may not meet the accuracy requirements of the antenna,which will lead to the appearance of uncertain factors caused by processing and manufacturing errors.The existence of these uncertain factors may lead to the deviation of the electrical performance of the working frequency band of the antenna,and even make the antenna not meet the use requirements.In order to solve the above problems,this thesis studies two different kinds of antenna robust optimization methods,including a parallel single-objective robust optimization algorithm and a multi-objective robust optimization algorithm.The first type of robust optimization method is based on known machining tolerances and the main purpose is reducing the response fluctuation of the design target.The other type of robust optimization method is based on unknown machining tolerances and the purpose is to obtain certain machining tolerances.This thesis is organized as follows:(1)The mathematical models of two kinds of antenna robust optimization are established.This chapter describes the problems of antenna robust optimization.This chapter constructs two kinds of different robust optimization model,and describes the construction methods of two kinds of robust optimization mathematical models in detail.The first construction method of mathematical model is based on 6 sigma quality management engineering,and the second is based on interval uncertainty analysis.Numerical examples are used to verify the two methods and the results show the feasibility in this thesis.(2)The single-objective robust optimization based on parallel meta-model solving algorithm of the antenna is proposed.This parallel robust optimization algorithm based on adaptive meta-model integrates the 6 sigma optimization model,GP model,and parallel BO algorithm and the algorithm flow is introduced in detail.Compared with the EGO robust optimization algorithm,two numerical examples and two ultra-wideband antennas are used to verify by proposed algorithm.The results show that the algorithm in this thesis has good accuracy and efficiency compared with the existing algorithms.(3)The multi-objective robust optimization of the antenna based on meta-model solving algorithm is proposed.This robust optimization algorithm based on adaptive meta-model integrates the interval analysis method,GP model,and NSGA-II algorithm and the algorithm flow is introduced in detail.A numerical example and two broadband antennas with low-profile are used to verify the method,and the antennas are processed to compare the electromagnetic performance with the simulation,and the results shows the accuracy and effectiveness of the proposed method.(4)The software of antenna robust optimization design is developed using the interface of the C++ Builder and MATLAB.The C++ Builder is used for the development of the main interface,and MATLAB is used to write the proposed two antenna robust optimization algorithms.Through the robust optimization of two wideband antenna examples,the feasibility and effectiveness of the software in practical engineering are proved.