遗传算法在半导体器件参数提取软件中的实现

随着半导体器件参数的增加，目标函数的自变量空间维数变得越来越大。传统的优化算法已经不能很好地处理此类问题，主要表现在无法有效地达到目标函数的极小点。文章首先在参数提取软件中实现了遗传算法，并与传统的优化算法进行了比较，探讨了该算法在参数提取软件中的实用性，提出了将遗传算法与传统优化算法结合在一起的方法。

Implementation of Genetic Algorithm in Semiconductor Device Parameter Extraction System

With the increase of the number of parameters in semiconductor devices, the dimensions of space in which the object function exists become larger and larger. At thesame time, the traditional optimization algorithms based on derivatives are getting incapable in solving this kind of problem. In this paper, the genetic opti mization algorithm is implemented in device parameter extraction system and compared with the traditional algorithms. The practicability of the genetic algorithm in device parameter extraction system is discussed, and a new method to combine the two algorithms is proposed.