›› 2012, Vol. 25 ›› Issue (1): 101-.

• 论文 • 上一篇    下一篇

基于可分解的多目标进化算法的PDN阻抗的优化

赵志超,张申,张辉   

  1. (1.西安电子科技大学 电子工程学院,陕西 西安 710071;2.西安电子科技大学 天线与微波技术国家重点实验室,陕西 西安 710071)
  • 出版日期:2012-01-15 发布日期:2012-01-10
  • 作者简介:赵志超(1986—),男,硕士研究生。研究方向:电路与系统。张申(1987—),女,硕士研究生。研究方向:电磁场与微波技术。

Optimization of PDN Impedance for the Multiobjective Evolutionary Algorithm Based on Decomposition

 ZHAO Zhi-Chao, ZHANG Shen, ZHANG Hui   

  1. (1.School of Electronic Engineering,Xidian University,Xi'an 710071,China;
    2.National Key Laboratory of Antennas and Microwave Technology,Xidian University,Xi'an 710071,China)
  • Online:2012-01-15 Published:2012-01-10

摘要:

频率的电源分配网络设计是一个多目标优化问题。应用多目标进化算法优化电源分配网络的阻抗,其中去耦电容的个数和种类成为了PDN中两个需要优化的目标函数,这使得PDN中的输入阻抗,在截止频率内小于目标阻抗以达到设计要求。为解决这个问题,应用可分解的多目标进化算法,同时优化这两个目标,以获得期望的Pareto Front(PF)。实验证明,该设计方法易于实现,且效果良好、稳定性强。优化的PDN的输入阻抗满足设计要求并且优化的去耦电容的个数和种类逼近PF。

关键词: PND, 去耦电容器, MOEA/D, 目标阻抗, 截止频率, PF

Abstract:

The frequent-domain design of a power distribution network (PDN) is a multi-objective problem (MOP).In this paper the multi-objective evolutionary algorithm (MOEA) is used to optimize power delivery network impedance,and the number and type of the decoupling capacitor become objective functions for optimization.This approach makes the input impedance of the PDN less than the target impedance within the cut-off frequency to meet design requirement.In order to solve this problem,the algorithm of MOEA/D is adopted in the simulation to optimize the two objectives simultaneously,and then we get the desired Pareto front (PF).Simulation results show that our design method is feasible for practical applications,has good stability and can achieve excellent results and that the Pareto Frontier we get meet the design requirement.Moreover,the number and type of the decoupling capacitor optimized is close to PF.

Key words: PDN;decoupling capacitor;MOEA/D;target impedance;cut-off frequency;PF

中图分类号: 

  • TM761