Resonant Frequency Modeling of Microwave Antennas Using Gaussian Process Based on Semisupervised Learning

Abstract

For the optimal design of electromagnetic devices, it is the most time consuming to obtain the training samples from full wave electromagnetic simulation software, including HFSS, CST, and IE3D. Traditional machine learning methods usually use only labeled samples or unlabeled samples, but in practical problems, labeled samples and unlabeled samples coexist, and the acquisition cost of labeled samples is relatively high. This paper proposes a semisupervised learning Gaussian Process (GP), which combines unlabeled samples to improve the accuracy of the GP model and reduce the number of labeled training samples required. The proposed GP model consists two parts: initial training and self-training. In the process of initial training, a small number of labeled samples obtained by full wave electromagnetic simulation are used for training the initial GP model. Afterwards, the trained GP model is copied to another GP model in the process of self-training, and then the two GP models will update after crosstraining with different unlabeled samples. Using the same test samples for testing and updating, a model with a smaller error will replace another. Repeat the self-training process until a predefined stopping criterion is met. Four different benchmark functions and resonant frequency modeling problems of three different microstrip antennas are used to evaluate the effectiveness of the GP model. The results show that the proposed GP model has a good fitting effectiveness on benchmark functions. For microstrip antennas resonant frequency modeling problems, in the case of using the same labeled samples, its predictive ability is better than that of the traditional supervised GP model.

Resonant Frequency Modeling of Microstrip Antenna Based on Deep Kernel Learning

“…Therefore, using a surrogate model instead of electromagnetic simulation software to evaluate the fitness of electromagnetic components can save optimization time, which is a popular topic of electromagnetic optimization design at present. Many modeling methods have been proposed by researchers, such as artificial neural network (ANN) [3][4][5], support vector machine (SVM) [6,7], extreme learning machine (ELM) [8][9][10], Gaussian process (GP) [11][12][13].…”

Antenna Modeling Based on Two-Stage Gaussian Process Considering Sensitivity Information

“…In contrast, GP has a strict statistical theoretical basis, which is suitable for dealing with small samples, high dimensions, nonlinear and other complex problems [24]. The strong modeling capabilities of GP make it possible to adaptively obtain hyper-parameters and realize probabilistic prediction that is different from other regression models, so it is more and more widely used in the analysis of antenna modeling problems [25], [26], [27], [28], [29]. When electromagnetic performance is evaluated and calculated by electromagnetic simulation software, if HFSS or CST software is used, the simulation results not only include conventional electromagnetic performance, but also provide sensitivity information [30], [31], [32], [33].…”

Antenna Optimization Design Based on Deep Gaussian Process Model

“…erefore, many literatures have proposed that artificial neural networks (ANNs) [2], support vector machine (SVMs) [3], and Gaussian process (GP) [4,5] can be used to analyze antenna problems. ANN can implement parallel processing, selflearning, and nonlinear mapping, but its structure is relatively complicated, which requires a large amount of electromagnetic simulation data, and it is difficult to determine with poor generalization ability [6].…”