A Multi-objective Optimization Method For Microwave Devices

In many optimization problems,multiple objectives are usually considered at the same time.However,the trade-off between objectives will lead to the failure of finding a unique optimal solution.Therefore,for multi-objective optimization problems,how to find an appropriate Parato optimal set has become a research focus.In microwave engineering field,with the increasing demand for higher frequency and higher speed,microwave devices become more complex.At the same time,more design specifications have to be met.It is of great practical significance to study the multi-objective optimization methods applicable to microwave devices.In this thesis,several existing multi-objective optimization algorithms are compared.The advantages and disadvantages of these methods analyzed.It usually takes a large number of objective function evaluations to obtain satisfactory results.Design optimization of microwave devices usually requires computationally expensive full-wave electromagnetics simulation.It is prohibitive to evalute objective functions excessively.The author proposes an efficient many-objective sequential domain patch optimization method to solve the two-and three-objective problems.The proposed algorithm is verified by test functions compared with the multi-objective genetic algorithm in terms of the computational cost and approximate Pareto front distribution.The results show that the proposed method is superior to the multi-objective genetic algorithm in terms of optimization efficiency.Finally,the proposed method is combined with the space mapping technology to optimize a microwave UWB antenna with multiple objectives.The result demonstrates the proposed method is suitable for microwave component design.