节点文献

基于神经网络的RFID标签天线印刷品质的优化研究

Optimized Research on Print Quality of RFID Label Antenna Based on Neural Network Model

【作者】 刘彩凤

【导师】 刘真;

【作者基本信息】 南京林业大学 , 制浆造纸工程, 2009, 博士

【摘要】 国内外研究表明:用导电油墨丝网印刷工艺制造RFID标签天线,是实现其小尺寸、低成本、大规模应用最具发展前景的方法。本文针对标签天线印刷品质优化控制中的评价指标、印刷参数优化设置、神经网络模型建立、天线结构尺寸预测与设计方案评价等一系列问题研究,取得了若干有价值的研究成果,主要创新点如下:①提出了RFID标签天线印刷品质的四个关键判据,并根据这四个关键判据,建立了印刷品质定量评价数学模型。这些关键判据及评价模型适合于包括射频标签在内的各种微电子产品印刷品质分析和评价。②创建了基于印刷品质优化的BP和RBF神经网络模型,使天线设计者可预先掌握最佳印刷品质状态下的导线墨膜实际尺寸,并可根据该实际尺寸修正标签天线的各项电性能指标,从而可实现天线各项电性能指标与印刷参数的综合优化。③通过神经网络模型仿真预测,可向印刷人员推荐最合适的丝网网版及最佳刮印压力、刮印速度等一组印刷参数,可取代过去采用反复上机调节丝网网版分辨率及印刷参数的传统方法。实现了印刷过程的误差可控,达到高效率、低成本和优品质的目的。总之,本文研究成果使电子专家和印刷专家密切合作,可确保设计天线和印制天线的性能一致。

【Abstract】 Research at home and abroad show that, RFID tag antenna fabricated through conductive ink screen printing process is the most promising approach of minisize, low cost, large-scale application.In this paper, printing labels antenna in the optimal control of the quality of evaluation, the printing parameter optimization settings, the establishment of neural network model, the antenna size and design of the forecast range of issues such as program evaluation studies were all discussed. Some research results have been obtained and the main innovation results are as follows:①Four key criteria of antenna RFID label printing quality were proposed, and quantitative evaluation of the quality of printing mathematical model was established according to these four key criteria. These key criteria and the evaluation model are suitable for analysis and evaluation of print quality of a variety of microelectronic products including RFID tags.②Both BP and the RBF neural network model based on optimized print quality was established, so that antenna designers can to control the actual conductor ink membrane size under the best quality printing state, and amend all kind of electrical performance indicators of the tag antenna according to the actual size, so that the electrical performance indicators of the tag antenna and print parameters can be optimize synthetically.③Through the neural network model for simulation of the forecast, a set of printing parameters including the most a appropriate silk screen, scrape print pressure and speed and so on, can be recommended the printing staff, Which replace the traditional methods that repeatedly adjust the screen resolution and printing parameters on machine in the past. The error controllability, high efficiency, low cost and high quality during printing process can all be obtained.In short, the results of this study can help cooperate between electronic experts and print specialists ,and ensure the performance accordance of the design antennas and print antenna.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络