CAJ | 학술논문

提出了一种多目标形式的空间映射算法，并将其应用于微带滤波器的优化设计问题。该算法没有采用ADS仿真软件进行粗糙模型的计算，而是采用多目标粒子群算法结合矩阵数值计算的形式，优化计算微带线的S参数，并应用神经网络映射粗糙模型与精确模型的仿真结果，由粗糙模型和神经网络推测出精确模型的帕累托最优解集。
제출료일충다목표형식적공간영사산법，병장기응용우미대려파기적우화설계문제。해산법몰유채용ADS방진연건진행조조모형적계산，이시채용다목표입자군산법결합구진수치계산적형식，우화계산미대선적S삼수，병응용신경망락영사조조모형여정학모형적방진결과，유조조모형화신경망락추측출정학모형적파루탁최우해집。
Multi-Objective Space-Mapping is proposed .Furthermore ,it is applied to the optimal design of microstrip filter . Instead of using ADS software ,a method combining Multi-Objective particle swarm optimization with the Matrix theory is used to simulate S parameters of coarse model ,and then artificial neural network (ANN ) is used to map the coarse model to fine model . Together with coarse model and ANN ,Fine model’s Pareto optimal set can be achieved .