Application of the Imanishism-based genetic algorithm to the module placement problem

摘要

In this paper, a new Genetic Algorithm(GA) based on evolutionary theory advocated by Kinji Imanisi is applied to the practical module placement problem in which the actual wire length among modules is estimated by the Semiperimeter wire length considering the difference of netlist. In order to reduce the computational time of GA, a new technique named Improved Harmonic Crossover (IHC) is also proposed. Finally, the performance of the proposed IHC is compared with the conventional Genetic Local Search method.