Microwave Filter Aided Design Method Based On Space Mapping

With the rapid development of modern communication industry,design requirements for microwave devices such as filters have become higher and higher,and the complexity of design has become increasingly high.Now,a large number of CAD tools such as Ansoft HFSS and Agilent ADS can be used to design microwave products.These tools can not only design microwave circuits,but also promote circuit optimization to improve power,size,timing,and other performance.However,the three-dimensional electromagnetic simulation software usually takes a long time to simulate a model,which reduces the designer's efficiency.Fortunately,the proposal of the space mapping method solved this problem to a large extent before long.This paper mainly focus on the application of space mapping method in filter design.In this study,the original space mapping algorithm(OSM),aggressive space mapping algorithm(ASM)and implicit space mapping algorithm(ISM)were studied and compared in detail.The advantages and disadvantages of the three space mapping algorithms were given,and an improved aggressive space was proposed.The parameter extraction is the most important step in the process of space mapping.This paper first reviewed the advantages and disadvantages of common parameter extraction algorithms(such as Cauchy method,vector fitting method and direct optimization method,etc.),and analyzed the error and constraints of parameter extraction.The L-BFGS-B algorithm was used as the optimization algorithm in this work.The coupling matrix was used as the optimization variable.The appropriate constraints were added to the optimization variables.The coupling matrix synthesis method and Cauchy method were adopted together to calculate the initial value of the coupling matrix.The only standard for the objective function was determined by the simultaneous consistency of coupling matrix's obtaining S-parameters and the filter response.In terms of space mapping algorithm iteration,this paper applied Python and HFSS to achieve the automation of space mapping algorithm.Further,this work described the development of filter-aided design software based on space mapping algorithm in detail.Firstly we introduced the advantages of Python in software development,and gave a variety of current methods to call HFSS and HFSS programming.Subsequently,the development of a Python custom development kit(Py Hfss)was proposed based on the HFSS COM interface.The use of Py Hfss simplified the operational flows of HFSS,making it easier to implement space mapping algorithm iterations and to improve design efficiency.Further,the development framework of our software was put forward based on Qt.The modules of our software and the way to use the software were all introduced.Finally,taking the 4th-order cross-coupling filter as an example,we gave the setting methods for coarse and fine model constraints,step length setting and convergence conditions of the space mapping algorithm.