泛研网

公众版 查看说明
全球科研项目数据库2026版

面向未来无线可重构智能射频模块与神经网络建模研究

项目来源

国家自然科学基金(NSFC)

项目主持人

南敬昌

项目受资助机构

辽宁工程技术大学

项目编号

61971210

立项年度

2019

立项时间

未公开

研究期限

未知 / 未知

项目级别

国家级

受资助金额

65.00万元

学科

信息科学-电子学与信息系统-电路与系统

学科代码

F-F01-F0118

基金类别

面上项目

关键词

智能系统 ; 可重构 ; 神经网络 ; 逆向建模 ; 射频电路系统 ;

参与者

高明明；李蕾；张沛泓；刘超；徐维；陶成健；刘婧；臧净；王宛

参与机构AI

辽宁工程技术大学

项目标书摘要：为实现未来无线射频微波前端系统的小型化和智能化，提高射频模块的设计精度和效率以及用于设计前端模块和构建智能系统的射频微波CAD软件的精确性、智能性和快速性，提出面向未来无线通信的可重构智能模块与神经网络智能建模研究。.本项目通过对可重构理论的研究，从可重构器件入手，进行多波段可重构功率放大器、通带内连续可调滤波器和陷波可重构超宽带天线等射频模块的设计，探索可重构无线系统的智能化机理、结构和方法，开发通用可重构射频电路模块，为构建智能化系统提供理论、方法和产品支持。依据人工智能神经网络和非线性散射参数(X参数)等理论，通过实验仿真和分析计算，研究射频模块神经网络(逆向)建模方法、适用结构及其算法和基于测量X参数建模方法，揭示建模的机理和规律，形成利用人工智能进行射频(集成)电路模块设计的方案与理论，最终提高建模的速度和精度，使模块和系统设计更精确化、便捷化和智能化，为电路软件提供方案。

Application Abstract: To realize the miniaturization and intelligence of the future wireless RF microwave front-end systems,improve the accuracy and efficiency of the RF modules design,and enhance the accuracy,intelligence and rapidity of the RF microwave CAD software,which is used for designing the front-end modules and building the intelligent systems.Intelligent re-configurable modules and intelligent neural network modelings for future wireless communication are proposed..By the research of re-configurable theories,this project will start with re-configurable components and design RF modules such as multi-band re-configurable power amplifiers,filters with continuously adjustable characteristics at pass-band,and ultra wide-band re-configurable antennas with band-notched characteristics,etc.We will explore the intelligent mechanisms,structures and methods of re-configurable wireless systems,develop universal re-configurable RF circuit modules to provide theories,methods and product support for intelligent systems.According to the theories of artificial intelligence neural network and nonlinear scattering parameters(X parameters),we will study RF module neural network(reverse)modeling methods,application structures and algorithms,and X-parameter modeling methods based on measurements by simulating and analyzing to reveal the mechanisms and laws of modeling,and form schemes and theories for the design of RF(integrated)circuit modules using artificial intelligence.Finally,the speed and accuracy of modeling will be improved by the researchful theories,methods and solutions,and make the design of modules and systems more precise,convenient and intelligent,and provide methods for the circuit design software.

项目受资助省

辽宁省

项目结题报告(全文)

本项目通过研究可重构理论、可重构智能模块与无源网络，为构建面向未来无线通信的智能无线系统提供了理论、技术和方法支持。主要研究并设计了双波段和多波段可重构功率放大器、通带内连续可调滤波器和陷波可重构超宽带天线等多款可重构射频模块，形成了一系列可重构宽带高效模块设计方法和理论。.研究了通用射频功率放大器各类设计方法。重点考虑了输出匹配网络、偏置网络和匹配结构，通过采用微带线结构、切比雪夫低通网络、含闭式解的宽带带通网络等方法设计输入输出匹配网络，并将具有滤波特性的电路匹配到输出电路中，以解决功放杂波影响的问题和偏置电路对功放性能影响的问题，增加功放设计的灵活性，从而实现功放模块的智能化。此外，项目还研究了开关类功放、宽带高效功放、Doherty功放以及双频功放等的设计方法。.研究了小型、宽带、高隔离度、可重构的不同类型天线和滤波器。针对现有超宽带天线和滤波器中存在的实际应用问题，利用缝隙结构、添加寄生枝节和加载缺陷地等技术方法实现了天线与滤波器的小型化、宽带化。研究设计的应用于不同系统天线和滤波器包括小型化超宽带天线、多频天线、陷波MIMO天线、新型多模谐振器的陷波超宽带小型化滤波器、双频带通滤波器和SIW滤波器等。.该项目还重点研究了射频模块神经网络建模方法和算法，提出或改进了多个智能模型和智能算法，为射频模块设计和智能系统提供了理论和技术支持。针对射频器件在计算量和设计效率上存在的不足，采用神经网络建模方式对相关模块进行建模，提出基于神经网络模型和基于深度学习网络架构的BP神经网络天线建模和新型深度多层感知机模型等多种算法模型，其中包括改进的粒子群算法、改进的蝴蝶算法等十余种算法模型，解决了算法陷入局部最优和过拟合现象，提高了射频建模精度、收敛速度并且降低了模型复杂度。此外针对在支撑集未知的情况下重构多频带信号的问题，提出基于压缩采样结构的盲多带信号预失真模型、基于稀疏假设下双频功率放大器的预失真模型等预失真算法。

  • 排序方式：
  • 1
  • /

  • 1.Design, fabrication and applications of flexible RFID antennas based on printed electronic materials and technologies

    • 关键词：
    • SILVER NANOPARTICLE INK; CONDUCTIVE INK; PAPER; TAG; FEATURES;DECOMPOSITION; FORMULATION; SUBSTRATE; NANOWIRES; CHEMISTRY

    The advancement of printed flexible electronics technology endows RFID tags with a simple fabrication process and innovative features such as flexibility, stretchability, and wearability. Flexible RFID tags based on printed flexible electronics technology are attracting considerable research interest. There are two main factors to consider when employing flexible printed electronics technology to create a flexible RFID tag antenna. One is the antenna structure design and the selection of conductive and flexible substrate materials, and the other is the fabrication of antenna prototypes and performance assessment that can satisfy specific demands. Both bring significant challenges to the materials, printing process, structural design and performance tuning of the antenna. Currently, most research is focused on the design and structure optimization of flexible RFID antennas, with just a few exceptions in inkjet printing based flexible electronics technology. Flexible antennas and printed flexible electronics technologies are yet to be efficiently combined, and systematic research and mutual development in terms of material selection, antenna design and simulation, prototype fabrication and measurement are yet to be achieved. Therefore, there is a strong need to do the research work in this field. This paper summarizes the current advances in flexible printed RFID antennas, with particular attention to antenna design, material selection, fabrication technologies, and applications. The challenges in each aspect are also explored. Further research should concentrate on the development of innovative materials with improved electrical properties and good surface functionality, the optimization of the fabrication processes, and the intelligent structural design of RFID antennas for multifunction operation. This paper is expected to provide guidance and new ideas for the development of flexible RFID tag antennas.

    ...
  • 排序方式：
  • 1
  • /