无线电通信技术2025，Vol.51Issue(2)：407-418,12.DOI:10.3969/j.issn.1003-3114.2025.02.024
天线和射频智能设计技术
Antenna and RF Rntelligent Design Technology
摘要
Abstract
In recent years,thanks to the rapid development of AI,machine learning,and optimization algorithms,significant pro-gress has been made in antenna and radio frequency design.These technologies can significantly improve engineers'design efficiency,especially in areas such as multi-parameter and multi-objective antenna optimization,as well as irregular antenna and radio frequency structures.AI-based design methods can accurately predict antenna performance parameters,optimize geometrical structures,and quickly generate innovative design solutions.In addition,intelligent algorithms are widely used in radio frequency circuit parameter op-timization,nonlinear effect calibration,and performance enhancement of complex multiple input multiple output systems.The introduc-tion of automation tools has significantly shortened the cycle from conceptual design to verification,while reducing R&D costs.This pa-per summarizes common techniques and methods in radio frequency intelligent design,using an automated optimization tool based on Gaussian regression models as an example to verify its application in base station antenna design.Through just 38 full-wave simula-tions,a design solution was successfully optimized to achieve a voltage standing wave ratio of less than 1.5,cross-polarization isolation greater than 15 dB,and a synthetic gain greater than 9 dBi within the 2.48～2.72 GHz frequency range,with the entire optimization process fully automated.Compared to traditional optimization methods,the single-fidelity optimization algorithm saved more than one-third of the time;further,multi-fidelity model-based optimization improved efficiency,saving an additional one-fifth of the time.To fa-cilitate quick comparisons,the models used in the study were pre-optimized by engineers.If original un-optimized models were used,the time savings from optimization would be even greater.These results fully demonstrate the application potential and technical feasi-bility of intelligent design in antenna and radio frequency optimization.关键词
智能设计/机器学习/高斯回归/天线优化/射频优化Key words
intelligent design/machine learning/Gaussian regression/antenna optimization/radio frequency optimization分类
电子信息工程引用本文复制引用
田珅,尹卫爽,方轶圣..天线和射频智能设计技术[J].无线电通信技术,2025,51(2):407-418,12.基金项目
国家重点研发计划(2023YFB2906104) National Key R&D Program of China(2023YFB2906104) （2023YFB2906104）
