基于改进QPSO算法的双陷波超宽带天线建模  

Modeling of dual-notch UWB antenna based on improved QPSO algorithm

作　　者：刘文进[1] 许馨水 南敬昌[1] 高明明[1] LIU Wenjin;XU Xinshui;NAN Jingchang;GAO Mingming(School of Electronics and Information Engineering,Liaoning Technical University,Huludao 125105,China)

机构地区：[1]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105

出　　处：《传感器与微系统》2022年第10期13-17,共5页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(61971210);国家自然科学基金青年科学基金资助项目(61701211);辽宁省特聘教授项目(551710007004)。

摘　　要：为了提高超宽带(UWB)天线的建模精度,提出一种基于改进的量子粒子群优化(QPSO)算法优化神经网络的建模方法。在QPSO算法中引入维数搜索策略,优化粒子组成,改善QPSO算法易陷入局部最优和全局收敛速度慢等问题;采用Elman神经网络作为基础神经网络,通过改变Elman神经网络的拓扑结构并引入自反馈增益因子,提高其泛化能力,用改进后的QPSO算法优化神经网络的权值阈值,提高模型的预测精度。将该模型用于一种UWB陷波的天线建模中,对天线的电参数进行仿真建模,实验结果表明:该建模方法平均绝对误差减小98.25%,运行时间减少34.81%,具有更高的预测精度和更快的收敛速度。In order to improve modeling precision of ultra-wideband(UWB)antenna, a neural network modeling method based on improved quantum particle swarm optimization(QPSO)is proposed.In this method, the dimension search strategy is introduced into the QPSO algorithm to optimize the composition of particles, so as to improve the local optimization and slow global convergence speed of QPSO algorithm.The Elman neural network is used as the basic neural network, and the topological structure of Elman neural network is changed and the self feedback gain factor is introduced to improve its generalization ability, the weight threshold of the neural network is optimized by the improved QPSO algorithm to improve the prediction precision of the model.The model is applied to the modeling of an UWB notch antenna, and the electrical parameters of the antenna are simulated.The experimental results show that the average absolute error of the modeling method is reduced by 98.25 %,and the running time is reduced by 34.81 %.It has higher prediction precision and faster convergence speed.

关 键 词：量子粒子群优化算法 ELMAN神经网络 维数搜索策略 超宽带天线 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]