Machine Learning-Assisted Ultra-Wideband Antenna Design Based on Heteroscedastic Evolutionary Bayesian Optimization

The development of wireless communication has raised higher requirements for antenna design and optimization. Traditional optimization methods are time-consuming. A machine learning-assisted antenna design method based on heteroscedasticity evolutionary Bayesian Optimization (HEBO) is proposed. A set of test functions and an ultra-wideband antenna are selected to verify the algorithm performance. The results show that compared with traditional optimization methods, the proposed method has better performance improvement in both test functions and antenna design.