A new implicit space mapping with high convergence for microwave filter design

A novel space mapping algorithm with high convergence is presented that improves the parameters mapping from surrogate model to fine model. By adding the process of coarse model parameter selection, it avoids false convergence in the optimization of surrogate model and speeds up the approximation between fine model and design object. No extra fine model evaluation is necessary in the parameter selection process, the optimization efficiency is improved. In this paper, a hairpin filter is designed and is compared with previous implicit space mapping algorithm, the results are better than the design specifications. The new algorithm is verified faster and more efficient.