Generalized Response Features

Solving simulation-driven design problems using response features has numerous advantages that include improved computational efficiency and reliability. These and other benefits have been extensively discussed in this book in the context of local and global parametric optimization, general-purpose surrogate modeling, multicriterial design, and uncertainty quantification. At the same time, it has been mentioned that the response feature technology comes with certain limitations. The major issues are related to the very definition, existence, and extractability of the characteristic points of the system outputs. On the one hand, the feature points are strongly related to the response shape. Consequently, their definition is problem-dependent and normally requires case-to-case extraction procedures. On the other hand, in some cases of severely distorted characteristics, the feature points may not exist, which entails applicability limitations. For this reason, the response feature technique is more suitable for problems such as local parameter tuning, statistical analysis, robust design, or performance-driven modeling, where the likelihood of degenerated responses is reasonably small. In any case, application of response feature requires customized approach, with the characteristic points individually defined for each type of the system outputs and tailored to a particular type of design specifications. This requires user experience and hinders a widespread application of FBO. This chapter discusses a generalized and unified feature point definition in the context of antenna design. It is suitable for majority of typical antenna input characteristics (narrow-, multiband, enhanced bandwidth, wideband) and performance specifications (matching improvement, bandwidth enhancement, mixture thereof). The presented framework allows for an automated definition of the feature points given the performance specifications, along with their extraction from EM-simulated responses. Its operation is illustrated using a range of planar antennas and favorably compared to conventional (nonfeature-based) design closure task formulation.