BP神经网络在异向介质基本结构分析中的应用

姜宇; 肖鸿; 刘兴鹏; 滕巍

doi:10.3724/SP.J.1146.2008.01703

BP神经网络在异向介质基本结构分析中的应用

doi: 10.3724/SP.J.1146.2008.01703

计量
- 文章访问数: 4005
- HTML全文浏览量: 225
- PDF下载量: 969
- 被引次数: 0
出版历程
- 收稿日期: 2008-12-15
- 修回日期: 2009-10-09
- 刊出日期: 2010-01-19

Applications of BP Neural Network in Analyzing Metamaterials Elemental Basic Structure

摘要

摘要: 为了减少传统数值分析法由于厚度谐振而引起的结果错误问题，实现异向介质高分析精度与高效率的共存，建立基于反向传播多层前馈型神经网络(BP神经网络)的异向介质电磁特性与介质敏感结构参数之间的神经网络模型，对异向介质的基本结构进行分析。实验结果表明，采用量化共轭梯度法的分析时间为145.535648 s，训练均方误差为0.00020679，所得结果与全波分析相吻合，满足工程要求。有效地克服了再次预测过程中提取参数的不稳定性，为异向介质的分析提供一种快速而准确的方法。
- 电磁特性参数; 异向介质; BP神经网络; 量化共轭梯度法
Abstract: To reduce the faults caused by thickness resonance in the traditional numerical analysis method, and achieve the coexistence between the high analysis precision and the high efficiency of the Left-Hand Materials(LHMs), the neural network model based on the back propagation multi-layer forward feed neural network is built between electromagnetism characteristic parameters and the dielectric sensitive structural parameters, which is used to analyze the basic Left-Hand Materials structural. The experimental results show that the analysis time is 145.535648 seconds and the training mean square error is 0.00020679 while adopting the scaled conjugate gradient method. The results achieved in the paper are coincident with these ones by the full wave method, satisfying the engineering demand. It effectively overcomes the instability while extracting the parameters in re-forecasting process, and provides a method with rapid and precise for the structure of Left-Handed Materials.

HTML全文

参考文献(1)

[1] 冉立新，陈红胜，皇甫江涛. 异向介质研究进展[J]. 微波学报，2004, 20(3): 89-95.Ran Li-xin, Chen Hong-sheng, and Huangfu Jiang-tao.Research progress for left-handed material[J]. Journal ofMicrowaves, 2004, 20(3): 89-95. [2] 潘攀，吴群，孟繁义. 一种微带左手传输线的精确分析方法[J].遥测遥控， 2007, 28(5): 20-25.Pan Pan, Wu Qun, and Meng Fan-yi. An approach foraccurate analysis of the microstrip left-handed transmissionline[J]. Journal of Telemetry, Tracking and Command, 2007,28(5): 20-25. [3] Smith D R, Padilla W J, and Vier D C. Composite mediumwith simultaneously negative permeability andpermittivity[J]. Physical Review, 2000, 84(18): 4184-4187. [4] 陈红胜. 异向介质等效电路理论及实验的研究[D]. [博士论文],国际电磁科学院浙江大学分院，2005.Chen Hong-sheng. Study of equivalent circuit theory andexperimentation for left-handed materials[D]. [Ph.D.dissertation], The Electromagnetics Academy at ZhejiangUniversity, 2005. [5] 吴群，武明峰. 基于传输线理论的SRR 结构异向介质的研究[J]. 电波科学学报，2006, 21(3): 310-314.Wu Qun and Wu Ming-feng. Research on SRR structuremetamaterial based on transmission line theory[J]. ChineseJournal of Radio Science, 2006, 21(3): 310-314. [6] 李蓉，王安国，吕英. 基于神经网络的新型缺陷接地结构优化设计[J]. 微波学报，2007, 23(9): 152-155.Li Rong, Wang An-guo, and Li Ying. Optimization design of anovel defected ground structure based on neural network[J].Journal of Microwaves, 2007, 23(9): 152-155. [7] 魏海坤. 神经网络结构设计的理论与方法[M]. 北京：国防工业出版社，2005: 4-6.Wei Hai-kun. The theory and method for designing the neuralstructure [M]. Beijing: GuoFang Industry Press, 2005: 4-6. [8] 刘永，张立毅. BP 和RBF 神经网络的实现及其性能比较[J].电子测量技术，2007, 30(4): 77-80.Liu Yong and Zhang Li-yi. Implementation of BP and RBFneural network and their performance comparison[J].Electronic Measurement Technology, 2007, 30(4): 77-80. [9] 杨国良，吴旷怀. 利用BP 神经网络反算沥青路面结构层弹性模量的研究[J].中山大学学报，2008, 47(5): 53-57.Yang Guo-liang and Wu Kuang-huai. Back calculation oflayered elastic moduli of asphalt pavement using BP neuralnetwork[J]. Acta Scientiarum Naturalium UniversitatisSunyatseni, 2008, 47(5): 53-57. [10] Moller, Fodslette Martin. Scaled conjugate-gradientalgorithm for fast supervised learning[J].Neural Networks.1993, 6(4):525-533 [11] 孙英广. 神经网络在径流预测模型研究中的应用及软件实现[D]. [硕士论文], 大连理工大学，2005.Sun Ying-guang. Forecasting reservior eischarge using ANNmodel and its software system[D]. [Master dissertation],Dalian University of Technology, 2005. [12] Gong Jian-qiang and Chu Qing-xin. Effective electromagneticparameters extraction method for 1D left-handed material[C].Proceedings of Asia-Pacific Microwave Conference, Bangkok,2007: 11-14.

施引文献

资源附件(0)

访问统计