This paper proposes a one‐dimensional convolutional autoencoders (1D‐CAE) surrogate‐based electromagnetic (EM) optimization technique exploiting particle swarm optimization (PSO) algorithm for microwave filter design. The 1D‐CAE is a flexible neural network structure that can be used as surrogate model to handle forward modeling problems as well as solving inverse problems, such as extracting the coupling matrix from S‐parameters. In this paper, the 1D‐CAE is proposed to handle forward modeling problem. On the one hand, the 1D‐CAE model is adopted to as the surrogate model to the EM simulation solver to speed up the PSO algorithm optimization process. On the other hand, the 1D‐CAE model can update the data set and network parameters online to continuously improve the prediction accuracy of the surrogate model during the PSO algorithm iterations. Compared with other neural network structures, the 1D‐CAE model proposed in this paper requires less data and converges faster in the same test environment. To demonstrate the validity of the proposed method, it was used to design a sixth‐order and an eighth‐order cross‐coupled filter. The design results shown that the proposed method is valid.
The tuning of microwave filter is important and complex. Extracting coupling matrix from given S-parameters is a core task for filter tuning. In this article, one-dimensional convolutional autoencoders (1D-CAEs) are proposed to extract coupling matrix from S-parameters of narrow-band cavity filter and apply this method to the computer-aided tuning process. The training of 1D-CAE model consists of two steps. First, in the encoding part, one-dimensional convolutional neural network (1D-CNN) with several convolution layers and pooling layers is used to extract the coupling matrix from the S-parameters during the microwave filters’ tuning procedure. Second, in the decoding part, several full connection layers are employed to reconstruct the S-parameters to ensure the accuracy of extraction. The S-parameters obtained by measurement or simulation exist with phase shift, so the influence of phase shift must be removed. The efficiency of the presented method in this article is validated by a sixth-order cross-coupled filter simulation model tuning example.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.