基于量子遗传算法的功率二极管物理模型参数提取研究

发布时间：2018-06-02 02:42

本文选题：功率二极管 + 物理模型　；参考：《西华大学》2015年硕士论文

【摘要】：功率半导体器件在能源高效转换、系统精准控制及新能源利用等方面有着得天独厚的优势,如何利用功率半导体器件组成的电力电子系统来提高能源的利用效率、减少单位电量的使用成本、降低工业的污染排放是电力电子研究领域亟待解决的重要课题。这就需要系统深入研究功率半导体器件的工作机理及其物理模型的关键参数,了解其工作时的动态特性,从而实现高效可靠利用电力电子系统的目标。本文以PIN型功率二极管作为研究对象,系统分析了PIN型功率二极管的物理结构及其工作机理,对其静态、动态特性及物理模型进行了详细分析。提出一种运用Saber软件和Matlab软件协同,将实验结果波形与仿真数据相比较,通过量子遗传优化算法提取功率二极管物理模型技术参数的方法。该方法选用能够体现功率二极管空间电荷区及载流子分布特性的动态过程作为研究对象,运用Saber软件调用Matlab软件,并通过它们之间的数据接口来传递仿真和优化所获得的参数数据。在Matlab软件中将初始载入的试验结果波形数据与包含待提取参数的二极管物理模型的仿真数据相比对,运用量子遗传算法优化待提取的参数值,并将优化所得参数值传入Saber软件中继续仿真,通过AIM语言编程实现这样一个优化循环进程,最终实现试验结果波形与仿真波形高度相似,从而得出需要的功率二极管模型参数。最后,为验证该参数提取方法的有效性,进行了验证实验,通过实验可知该方法所提取出的参数能够准确描述功率二极管的动态特性,证明该方法是有效可靠的。
[Abstract]:Power semiconductor devices have unique advantages in energy efficiency conversion, system precision control and new energy utilization. How to use power electronic system composed of power semiconductor devices to improve energy utilization efficiency, Reducing the cost of electricity per unit and reducing industrial pollution emissions are important issues to be solved in the field of power electronics research. Therefore, it is necessary to study the working mechanism of power semiconductor devices and the key parameters of their physical models, to understand the dynamic characteristics of power semiconductor devices, and to realize the goal of using power electronic systems efficiently and reliably. In this paper, PIN type power diode is taken as the research object. The physical structure and working mechanism of PIN type power diode are systematically analyzed, and its static and dynamic characteristics and physical model are analyzed in detail. This paper presents a method of extracting the technical parameters of power diode physical model by quantum genetic optimization (QGA) by using Saber software and Matlab software to compare the experimental waveform with the simulation data. In this method, the dynamic process which can reflect the space charge region and carrier distribution characteristics of power diode is selected as the research object, and the Matlab software is called by Saber software. The parameter data obtained by simulation and optimization are transmitted through the data interface between them. Compared with the simulation data of diode physical model containing the parameters to be extracted, quantum genetic algorithm is used to optimize the parameters to be extracted by comparing the initial loaded waveform data with the simulation data of diode physical model in Matlab software. The optimized parameters are transmitted to the Saber software for further simulation. The optimization cycle process is realized by programming in AIM language. Finally, the waveform of the test results is similar to the simulation waveform, and the required parameters of the power diode model are obtained. Finally, in order to verify the validity of the method, the experimental results show that the parameters extracted by this method can accurately describe the dynamic characteristics of the power diode, and prove that the method is effective and reliable.
【学位授予单位】：西华大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN313.4

【相似文献】