| To solve the problem that traditional filter design is low optimization efficiency, space mapping method combines circuit optimization quickness and electromagnetic simulation accuracy, improves filter design efficiency.In this paper, space mapping technology is analyzed and implicit space mapping (ISM) algorithm is analyzed and researched deeply. We analyze the similarities and differences of implicit space mapping algorithm and explicit space mapping algorithm. By researching the advantages and disadvantages of implicit space mapping algorithm, the parameter selection technology is put forward and applied to the implicit space mapping algorithm. A new algorithm is formed, named the parameter selection implicit space mapping algorithm (PSISM). The parameter selection process reacts fully in rough models. It doesn't need to increase the times of fine model calculation in each iteration. It has improved the mapping between the coarse model and fine model parameters and avoided that false convergence arises.According to the filter synthesis theory, a hairpin filter and a parallel coupling line filter are synthesized by chebyshev filter synthesis method, a mirror inhibit filter coupling matrix is synthesized using quasi elliptic function. The hairpin filter is optimized by PSISM and ISM algorithm, and the results of the two algorithms are compared. The parallel coupling line filter and cross coupling filter are optimized by PSISM method. The filters designed have all meet the design requirements. In the optimize process of the algorithm, coarse model is adopted ADS circuit model and fine model is calculated by Ansoft HFSS.According to the filters'success design, it's verified that with the advantages of high effective and stability, PSISM algorithm accelerates dramatically filter design efficiency. It saves much precious time in the optimization of microwave device design. Not only in filter designing, but it also plays a very important role in other microwave components design. |
PDF Full Text Request