Surrogate‐assisted optimization of 3D printed ceramic nonuniform nonplanar microstrip filter

Abstract

Herein, by using 3D printing technology and data‐driven surrogate model‐assisted optimization method, design of a ceramic material‐based nonuniform nonplanar microstrip filter is taken into the consideration in a computationally efficient and low‐cost manner. For this aim, 3D EM model of the proposed design had been used for generating training and validation data sets. Then commonly used state‐of‐the‐art regression algorithms had been used for creating accurate and fast surrogate models to create a mapping between inputs of the model and outputs of the unit element design. After that, Grey Wolf Optimization algorithm had been used for design optimization of a bandpass filter. Then, via the use of 3D printer, the optimally designed filter had been prototyped and its performance characteristics are measured. As a result, by using 3D printer technology, ceramic material, and the proposed method, design optimization of nonuniform nonplanar microstrip filter can be achieved in a computationally efficient way.

Data‐driven modeling of band‐pass filter for sub‐5G applications

“…In this section, by using the obtained CNN surrogate model, the optimization of an EC‐BPF will be realized. A similar approach was proposed in Mahouti et al 17 is taken into consideration, where a design optimization process using Bayesian optimization. The following function has been used as the cost function to assist the optimization process.…”

“…The average simulation time is around 65 s using a simulation setup: Intel(R) Core(TM) i5‐6200U CPU @ 2.30 GHz, with 8 GB of installed RAM. Here for a design optimization of EC‐BPF any kind of optimization techniques would need to run for 30 iteration of solutions which usually corresponded to 1500 function evaluation (such as differential evolutionary algorithm [DEA] 17 with population size of 50 and iteration of 30 [default values for DEA]). The total duration of such process would be around of 27 h (1500 × 65 [total number of function evaluation × average simulation duration in seconds]) using direct EM optimization approach, while by using the data‐driven surrogate model approach this would be around 18 h (1000 × 65, total number of samples training and test).…”

“…One of the most efficient techniques for achieving this problem is utilization of artificial intelligence (AI) based algorithms for forming data‐driven surrogate models of microwave components 13–28 . During the last decade, various AI‐based techniques have been extensively used for modeling of different microwave components such as: modeling of microwave transistors, 13–16 microstrip transmission lines, 17–19 microwave antenna, 20–24 and reflectarray designs 25–28 .…”

A novel near field radiation shaping technique by using data driven surrogate based optimization for nondestructive hyperthermia

“…A viable and efficient solution for this problem is the usage of data driven surrogate models 19–23 . Application of data driven surrogate models for repetitive optimization process is a well‐known technique which is being studied for many different type of microwave and RF research topics such as; modeling of RF small signal microwave transistors, 24–28 design and optimization of passive microwave stages such as filters, 29–42 Meta surfaces, 43–45 and microwave antenna designs 46–57 …”

“…A viable and efficient solution for this problem is the usage of data driven surrogate models. [19][20][21][22][23] Application of data driven surrogate models for repetitive optimization process is a well-known technique which is being studied for many different type of microwave and RF research topics such as; modeling of RF small signal microwave transistors, [24][25][26][27][28] design and optimization of passive microwave stages such as filters, [29][30][31][32][33][34][35][36][37][38][39][40][41][42] Meta surfaces, [43][44][45] and microwave antenna designs. [46][47][48][49][50][51][52][53][54][55][56][57] In this work, in order to achieve an efficient optimization process for focusing the EM energy into the tumor location, a data driven surrogate model of antenna array, which its data samples are generated using ARM method, is taken into the consideration.…”