基于耦合矩阵诊断的滤波器设计方法研究
发布时间：2018-10-21 19:47
【摘要】：微波滤波器是各类微波系统的重要选频器件。目前,滤波器的结构变得越来越复杂,运用已有的方法设计这类复杂结构的滤波器存在着较大的困难,这就制约了滤波器的进一步发展。因此,研究一种针对复杂结构滤波器的设计方法具有很强的现实意义。本文的研究是围绕着如何实现复杂结构滤波器的快速设计而展开的,对耦合矩阵的综合理论、滤波器的诊断技术、滤波器等效电路模型的建立,以及基于神经网络的空间映射方法作了深入的研究,提出了一种基于耦合矩阵诊断的滤波器设计方法。本文的主要研究内容如下:首先,系统的介绍了耦合矩阵的综合理论和滤波器的诊断技术。一方面,介绍了耦合矩阵综合的分析法和优化法,它们是设计耦合谐振腔滤波器的基础。另一方面,介绍了三种分析法诊断技术和一种优化法诊断技术,它们可以用来提取滤波器的耦合矩阵。其次,提出了一种建立滤波器等效电路模型的精确方法,该方法对传统的等效电路模型建立方法进行了改进,即加入了FIR(frequency-invariant reactance)元件。建立的等效电路模型可以用于滤波器的参数提取,从而确保了参数提取的准确性。本文运用了两个例子对提出的等效电路模型建立方法进行了详细的论证,证明了该方法的合理性。然后,提出了一种基于神经网络的空间映射方法,该方法可用于滤波器的设计,特别是在结构复杂的滤波器设计上显示出了巨大的优势。另外,针对如何选取样本点的问题提出了一种全新的样本点选取方法,该方法可用于神经网络建模时训练样本和测试样本的选取,确保了样本点选取的合理性,从而能够利用选取的样本得到正确的映射。本文通过举例验证了提出的样本点选取方法的合理性。最后,提出了一种基于耦合矩阵诊断的滤波器设计方法,该方法将滤波器的诊断技术与基于神经网络的空间映射相结合,用在了滤波器的设计上。提出的滤波器设计方法是根据实物测试的响应曲线进行设计的,避开了使用电磁仿真软件进行优化或者验证,这与以往的滤波器设计方法有着很大的区别。同时,该方法是一种普适的设计方法,尤其是在复杂结构的滤波器设计上显示出了巨大的优势。本文以SIW(substrate integrated waveguide)滤波器为例对提出的设计方法进行了充分的验证,证明了该设计方法的有效性。
[Abstract]:Microwave filter is an important frequency selective device in various microwave systems. At present, the structure of filter becomes more and more complex, and it is difficult to design the filter with existing methods, which restricts the further development of filter. Therefore, it is of great practical significance to study a design method for complex structure filters. The research of this paper is about how to realize the fast design of complex structure filter. The synthesis theory of coupling matrix, the diagnosis technology of filter, the establishment of equivalent circuit model of filter are discussed. A filter design method based on coupling matrix diagnosis is proposed. The main contents of this paper are as follows: firstly, the synthesis theory of coupling matrix and the diagnostic technique of filter are introduced systematically. On the one hand, the analysis method and optimization method of coupling matrix synthesis are introduced, which are the basis of designing coupling resonator filter. On the other hand, three analytical diagnostic techniques and an optimal diagnostic technique are introduced, which can be used to extract the coupling matrix of filters. Secondly, an accurate method to establish the equivalent circuit model of filter is proposed. The traditional equivalent circuit model is improved by adding FIR (frequency-invariant reactance) element. The established equivalent circuit model can be used to extract the parameters of the filter, thus ensuring the accuracy of parameter extraction. In this paper, two examples are used to demonstrate in detail the method of establishing equivalent circuit model, and the rationality of the method is proved. Then, a neural network-based spatial mapping method is proposed, which can be used in the design of filters, especially in the complex structure of the filter design shows great advantages. In addition, aiming at the problem of how to select sample points, a new method of selecting sample points is put forward. This method can be used to select training samples and test samples in neural network modeling, which ensures the rationality of sample points selection. Thus the correct mapping can be obtained by using the selected samples. In this paper, the rationality of the proposed method is verified by an example. Finally, a filter design method based on coupling matrix diagnosis is proposed, which combines the filter diagnosis technique with the spatial mapping based on neural network, and is used in the design of the filter. The proposed filter design method is designed according to the response curve of the physical test, avoiding the use of electromagnetic simulation software for optimization or verification, which is very different from the previous filter design method. At the same time, this method is a universal design method, especially in the complex structure of filter design shows great advantages. In this paper, the SIW (substrate integrated waveguide) filter is taken as an example to verify the proposed design method, and the effectiveness of the proposed design method is proved.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN713
本文编号：2286180
[Abstract]:Microwave filter is an important frequency selective device in various microwave systems. At present, the structure of filter becomes more and more complex, and it is difficult to design the filter with existing methods, which restricts the further development of filter. Therefore, it is of great practical significance to study a design method for complex structure filters. The research of this paper is about how to realize the fast design of complex structure filter. The synthesis theory of coupling matrix, the diagnosis technology of filter, the establishment of equivalent circuit model of filter are discussed. A filter design method based on coupling matrix diagnosis is proposed. The main contents of this paper are as follows: firstly, the synthesis theory of coupling matrix and the diagnostic technique of filter are introduced systematically. On the one hand, the analysis method and optimization method of coupling matrix synthesis are introduced, which are the basis of designing coupling resonator filter. On the other hand, three analytical diagnostic techniques and an optimal diagnostic technique are introduced, which can be used to extract the coupling matrix of filters. Secondly, an accurate method to establish the equivalent circuit model of filter is proposed. The traditional equivalent circuit model is improved by adding FIR (frequency-invariant reactance) element. The established equivalent circuit model can be used to extract the parameters of the filter, thus ensuring the accuracy of parameter extraction. In this paper, two examples are used to demonstrate in detail the method of establishing equivalent circuit model, and the rationality of the method is proved. Then, a neural network-based spatial mapping method is proposed, which can be used in the design of filters, especially in the complex structure of the filter design shows great advantages. In addition, aiming at the problem of how to select sample points, a new method of selecting sample points is put forward. This method can be used to select training samples and test samples in neural network modeling, which ensures the rationality of sample points selection. Thus the correct mapping can be obtained by using the selected samples. In this paper, the rationality of the proposed method is verified by an example. Finally, a filter design method based on coupling matrix diagnosis is proposed, which combines the filter diagnosis technique with the spatial mapping based on neural network, and is used in the design of the filter. The proposed filter design method is designed according to the response curve of the physical test, avoiding the use of electromagnetic simulation software for optimization or verification, which is very different from the previous filter design method. At the same time, this method is a universal design method, especially in the complex structure of filter design shows great advantages. In this paper, the SIW (substrate integrated waveguide) filter is taken as an example to verify the proposed design method, and the effectiveness of the proposed design method is proved.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN713
【参考文献】
相关硕士学位论文 前1条
1 彭杰;微波滤波器机辅设计与调试技术研究[D];西安电子科技大学;2010年
，本文编号：2286180
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2286180.html
