Space Mapping Modeling and Intelligent Optimization of Planar Direct Coupled Filter

In this paper, a modeling and intelligent optimization method based on space mapping is proposed to deliver speed and accuracy of the overall simulation design of multi-order planar direct coupled filters. The mapping model between the physical structure and the electrical performance of the filter is constructed by using neural network, and the optimization objective function is constructed by the reflection coefficient curve. On this basis, the final physical size of the filter is obtained by iteratively adjusting the input of the neural network through the proposed algorithm. Finally, an optimization design of a 5-order thin film planar direct coupled filter is taken as an example to verify the proposed method.