改进的隐式空间映射算法在LTCC滤波器优化设计中的应用研究

发布时间：2018-05-18 18:05

本文选题：LTCC + 隐式空间映射算法　；参考：《浙江工业大学》2015年硕士论文

【摘要】：人们对电子设备尤其是移动设备的小型化需求不断提高,使得低温共烧陶瓷(LTCC,Low Temperature Co-fired Ceramic)技术作为一种上佳的封装技术而倍受工程师们的青睐。低温共烧陶瓷技术是在多层陶瓷基板上埋置无源组件,经过叠压共烧成三维结构,从而实现器件的小型化。此技术已被广泛应用于制作巴伦、滤波器和天线等功能器件。本文主要研究LTCC带通滤波器的设计和优化,并应用改进的隐式空间映射算法较好的解决了LTCC三维结构电磁仿真耗时长和优化方向不确定的问题,其主要工作有:1、选择切比雪夫低通原型滤波器,经过频率变换、元件变换以及J/K倒置变换器,得到要设计的微波频段上建模的带通滤波器集总电路,再结合传输零点方法和传输线理论,设计出4层LTCC微带滤波器。2、对空间映射算法进行了较深入的研究,重点着眼于本文所采用的隐式空间映射算法,并针对其在参数提取过程中出现的参数非唯一性问题,采用频率参数提取的方法进行了较好的解决,并设计了平行耦合微带滤波器,以此说明改进算法的有效性。3、采用HFSS搭建算法中所需的精细模型(fine model),用ADS中的多层元件库来实现相对应的粗糙模型(coarse model),通过采用了频率参数提取的隐式空间映射算法对模型进行不断的迭代优化,最终得到符合设计要求的带通滤波器,并与隐式空间映射算法的效能进行了比较。最终频率隐式空间映射算法经过了5次迭代后收敛,而隐式空间映射算法经过了7次迭代后才收敛。所设计的带通滤波器外形尺寸为2.5 mm?2.0mm?0.9 mm,中心频率为2.45GHz,带宽为100MHz,可应用于蓝牙和无线局域网(WLAN)频段。仿真优化的结果证明频率隐式空间映射算法相比一般隐式空间映射算法,收敛更快,优化成本更低。且两种算法相比参数扫描方式,较好地解决了LTCC滤波器仿真优化耗时长和优化方向不明确的问题。
[Abstract]:Due to the increasing demand for miniaturization of electronic devices, especially mobile devices, LTCC low Temperature Co-fired Ceramic-based technology is favored by engineers as a kind of excellent packaging technology. Low-temperature co-fired ceramic technology is to bury passive components on multilayer ceramic substrates and to achieve the miniaturization of devices by laminating and co-firing into three-dimensional structures. This technology has been widely used in the manufacture of Barron, filters and antennas and other functional devices. This paper mainly studies the design and optimization of LTCC bandpass filter, and solves the problems of long time consuming and uncertain optimization direction of LTCC 3D structure electromagnetic simulation by using the improved implicit space mapping algorithm. The main work is to select Chebyshev low-pass prototype filter, through frequency conversion, component transformation and J- / K inversion converter, to obtain the lumped circuit of the bandpass filter which is to be modeled on the microwave frequency band. Combined with the transmission zero method and transmission line theory, a 4-layer LTCC microstrip filter .2is designed, and the spatial mapping algorithm is studied deeply, focusing on the implicit spatial mapping algorithm used in this paper. Aiming at the non-uniqueness of parameters in the process of parameter extraction, the method of frequency parameter extraction is used to solve the problem, and a parallel coupled microstrip filter is designed. This shows the effectiveness of the improved algorithm. 3. The fine model needed in the algorithm is built with HFSS, and the corresponding coarse model is implemented by using the multilayer component library in ADS. The implicit spatial mapping based on frequency parameter extraction is adopted. The algorithm performs iterative optimization of the model. Finally, the bandpass filter which meets the design requirements is obtained, and the performance of the implicit space mapping algorithm is compared. The final frequency implicit space mapping algorithm converges after 5 iterations, while the implicit space mapping algorithm converges after 7 iterations. The designed bandpass filter has a size of 2.5 mm?2.0mm?0.9 mm, a central frequency of 2.45 GHz and a bandwidth of 100 MHz. It can be used in Bluetooth and WLAN (WLAN) band. The simulation results show that the frequency implicit spatial mapping algorithm has faster convergence and lower optimization cost than the general implicit spatial mapping algorithm. Compared with the parameter scanning method, the two algorithms can solve the problems of long time consuming and unclear optimization direction of LTCC filter simulation.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN713

【参考文献】