An age-controlled evolutionary algorithm for optimization problems in physical layout

We present the first evolutionary approach using an age-controlled population model for solving layout problems. In this approach the age is counted individually. Age-intrinsic information and possible conclusions are discussed elaborately. The age-intrinsic information is exploited to improve and accelerate the optimization. Selection and mutation benefit from this approach. Due to industrial needs the optimization characteristic is controlled by a single strategy variable /spl sigma/. This strategy is best-suited to discrete and combinatorial optimization problems. It facilitates efficient controlling of genetic operators. The presented strategy has been implemented and successfully applied to the assignment problem on analog transistor arrays. This paper focuses on the general problem-independent aspects.