稀疏正则化逆向神经网络在双陷波超宽带天线设计中的应用  

Application of sparse regularized inverse neural network in design of dual-notch UWB antenna

作　　者：南敬昌[1] 王梓琦 高明明[1] Nan Jingchang;Wang Ziqi;Gao Mingming(College of Electrics & Information Engineering,Liaoning Technical University,Huludao Liaoning 125105,China)

机构地区：[1]辽宁工程技术大学电子与信息工程学院

出　　处：《计算机应用研究》2019年第8期2473-2477,2482,共6页Application Research of Computers

基　　金：国家自然科学基金面上项目(61372058);辽宁省高校重点实验室项目(LJZS007);辽宁省教育厅科学研究一般项目(L2015209)

摘　　要：在双陷波超宽带天线的设计过程中,直接逆向神经网络模型精度较低,而BP逆向神经网络泛化能力较差,若单独使用HFSS仿真软件需要不断优化天线各参数增加了设计时间。针对以上问题,提出一种将HFSS与稀疏正则化逆向神经网络联合的方法。该方法在逆向神经网络性能函数中增加l1/2范数和l2范数,l1/2范数引入了新的权系数,扩充了输入样本向量,使网络更易得到稀疏性解,逆模型精度更高,l2范数能有效避免过拟合现象,使网络泛化能力更强。应用于双陷波超宽带天线设计中,采用在辐射贴片上开弧形槽的方式产生陷波特性,根据天线目标电压驻波比逆向求解对应的开槽尺寸。仿真实验结果表明,与BP逆向神经网络方法相比,求得的与天线电压驻波比对应的开槽角度相对误差减小了69.3%,开槽半径相对误差减小了88.7%,网络运行时间减少了15.9%;最终设计的天线带宽为2.4~11GHz,实现了3.31~3.8GHz和4.98~6.05GHz的良好陷波特性,缩短了整个天线的设计周期。In the design of dula band-notched ultra-wide band antennas,the direct inverse neural network model has lower precision,the BP inverse neural network has poor generalization ability,and if only use the HFSS simulation software,it is necessary to continuously optimize the antenna parameters which increasing the design time.Aiming at the above problems,this paper proposed a method combining HFSS with sparse regularized inverse neural netwrok.This method added l 1/2 norm and l 2 norm in the performance function of the inverse neural network.The l 1/2 norm introduced a new weight coefficient,expanded the input sample vector,made the network to obtain the sparsity solution more easily,and the inverse model got higher accuracy.Meanwhile,the l 2 norm avoided the over-fitting phenomenon effectively and made the network generalization ability stronger.It applied to the design of dual band-notched ultra-wide band antenna,using of arc grooves on radiating patches generated notch characteristics,and according to the antenna target voltage standing wave ratio(VSWR) solved inversely the corresponding slot size.Simulation results show that the relative error of slot angle which corresponding to VSWR of the antenna reduced by 69.3%,and the relative error of slot radius lessened by 88.7%,and the network running time decreased by 15.9% compared with BP neural network method.The final designed antenna bandwidth is 2.4~11 GHz,achieves the good notch characteristics in 3.31~3.8 GHz and 4.98~6.05 GHz,and shortens the entire antenna design cycle.

关 键 词：逆向神经网络 HFSS l1/2范数 l2范数 双陷波超宽带天线 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]