Expedited antenna optimization with numerical derivatives and gradient change tracking

Abstract

Purpose The purpose of this study is to propose a framework for expedited antenna optimization with numerical derivatives involving gradient variation monitoring throughout the optimization run and demonstrate it using a benchmark set of real-world wideband antennas. A comprehensive analysis of the algorithm performance involving multiple starting points is provided. The optimization results are compared with a conventional trust-region (TR) procedure, as well as the state-of-the-art accelerated TR algorithms. Design/methodology/approach The proposed algorithm is a modification of the TR gradient-based algorithm with numerical derivatives in which a monitoring of changes of the system response gradients is performed throughout the algorithm run. The gradient variations between consecutive iterations are quantified by an appropriately developed metric. Upon detecting stable patterns for particular parameter sensitivities, the costly finite differentiation (FD)-based gradient updates are suppressed; hence, the overall number of full-wave electromagnetic (EM) simulations is significantly reduced. This leads to considerable computational savings without compromising the design quality. Findings Monitoring of the antenna response sensitivity variations during the optimization process enables to detect the parameters for which updating the gradient information is not necessary at every iteration. When incorporated into the TR gradient-search procedures, the approach permits reduction of the computational cost of the optimization process. The proposed technique is dedicated to expedite direct optimization of antenna structures, but it can also be applied to speed up surrogate-assisted tasks, especially solving sub-problems that involve performing numerous evaluations of coarse-discretization models. Research limitations/implications The introduced methodology opens up new possibilities for future developments of accelerated antenna optimization procedures. In particular, the presented routine can be combined with the previously reported techniques that involve replacing FD with the Broyden formula for directions that are satisfactorily well aligned with the most recent design relocation and/or performing FD in a sparse manner based on relative design relocation (with respect to the current search region) in consecutive algorithm iterations. Originality/value Benchmarking against a conventional TR procedure, as well as previously reported methods, confirms improved efficiency and reliability of the proposed approach. The applications of the framework include direct EM-driven design closure, along with surrogate-based optimization within variable-fidelity surrogate-assisted procedures. To the best of the authors’ knowledge, no comparable approach to antenna optimization has been reported elsewhere. Particularly, it surmounts established methodology by carrying out constant supervision of the antenna response gradient throughout successive algorithm iterations and using gathered observations to properly guide the optimization routine.

EM-driven size reduction and multi-criterial optimization of broadband circularly-polarized antennas using pareto front traversing and design extrapolation

“…Owing to the extrapolation techniques described above, the initial designs are normally of good quality so local tuning is sufficient. Here, it is realized using an accelerated version of the trust-region (TR) gradient-search proposed in 40 . The algorithm is briefly outlined here for the convenience of the reader.…”

“…This is realized using inverse modeling methods 38 and extrapolation. Furthermore, the low cost of the parameter tuning process is ensured by utilization of trust-region gradient search with sparse sensitivity updates 39 , 40 . The proposed technique is demonstrated using three examples of wide-slot CP structures.…”

Accelerated parameter tuning of antenna structures using inverse and feature‐based forward kriging surrogates

“…Using U, the parameter tuning problem can be formulated as a nonlinear minimization task of the form where x * is the optimum design to be found. The problem (2) may be subject to additional constraints, which can be handled explicitly or implicitly (e.g., using the penalty function approach 55 ).…”

Design specification management with automated decision-making for reliable optimization of miniaturized microwave components

“…Improving computational efficiency of simulation-based design procedures has been targeted by numerous research endeavours. These efforts focused on the development of strictly algorithmic approaches, both intrusive (e.g., gradient-based procedures accelerated by means of adjoint sensitivities 27 , 28 ), and non-intrusive (e.g., trust-region methods with sparse sensitivity updates 29 , 30 , as well as surrogate-based frameworks involving data-driven 7 , and physics-based metamodels 5 . Although approximation surrogates (kriging 31 , radial-basis functions 32 , support vector regression 33 , polynomial chaos expansion 34 , 35 , neural networks 36 , Gaussian process regression 37 , polynomial regression 38 ) are by far more popular, their application is limited by the curse of dimensionality, which is particularly troublesome when handling nonlinear outputs of high-frequency structures.…”