Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (26): 53-57.DOI: 10.3778/j.issn.1002-8331.2010.26.018

• 研发、设计、测试 • Previous Articles     Next Articles

Multi-objective optimization and evolutionary design of analog circuits

XIA Xue-wen1,XIONG Zeng-gang1,LI Yuan-xiang2,ZHU Ji-xiang2   

  1. 1.School of Computer and Information Science,Xiaogan College,Xiaogan,Hubei 432000,China
    2.State Key Laboratory of Software Engineering,Wuhan University,Wuhan 430072,China
  • Received:2009-09-21 Revised:2009-11-18 Online:2010-09-11 Published:2010-09-11
  • Contact: XIA Xue-wen



  1. 1.孝感学院 计算机与信息科学学院,湖北 孝感 432000
    2.武汉大学 软件工程国家重点实验室,武汉 430072
  • 通讯作者: 夏学文

Abstract: The definitions and details of multi-objective during process of analog circuit’s design are proposed in this paper.Based on it,a Multi-Objective Genetic Algorithm(MOGA) is presented.It features a group of uniformly scattered search direction by using Non-dominated Sorting Genetic Algorithms(NSGA) and fitness sharing strategy.A primitive circuit representation scheme,circuit-constructing instruction,is presented that enables circuit structure to be generated and improve the efficiency of evolution.And this representation is also the same with digital circuit.Coevolutionary fitness-evaluation strategy is utilized to improve the learning ability and efficiency of evolution.Some experimental results presented show that the MOGA is capable of searching out simple and practical circuits.

Key words: evolutionary design of circuit, non-dominated sorting genetic algorithms, multi-objective optimization, coevolutionary fitness-evaluation

摘要: 对模拟电路设计中涉及的多个目标进行了定义与量化,并针对这些目标提出一种面向模拟电路演化设计的多目标遗传算法,该方法利用非支配排序和适应值共享策略来提高搜索方向的空间均匀性,引入基于电路构造指令的编码方案来支持电路自动生成和提高电路演化的效率,并且该编码方案也同样适用于数字电路。利用协同演化的适应值评估策略来增强种群的学习能力,提高演化效率。实验结果表明,该方法可以设计出更实用、简单的模拟电路。

关键词: 演化电路设计, 非支配排序遗传算法, 多目标优化, 协同演化适应值评估

CLC Number: