Abstract: A novel multi-objective optimization based on Suitable-Distribution Particle Swarm（SDPS） is discussed in this paper.The elitism set is made up of non-dominated solutions with high Suitable-Distribution Value（SDV）.The suitable-distribution radius is proposed to guarantee the diversity of the individuals in the elitism set and make them well distributed in the objective space.SDPS is applied to SOC architectural parameters design and a triple discretization method is used to discretize the velocity vector of the particles.The experimental result indicates that,within the two dimension objective space composed of power consumption and executive time cost,the configurations gained by SDPS have well distribution and nondominance and SDPS is very efficient in SOC architectural parameters searching.

Key words: suitable-distribution, particle swarm, multi-objective optimization, SOC, architectural parameters design

摘要： 提出了一种基于适配粒子群的多目标优化方法,其精英集由搜索过程中适配值较高的非劣解构成。适配半径的提出有利于保持精英集中个体的差异性,使得解集能够在目标空间中均匀分布。将该多目标优化方法应用于SOC结构参数设计,并且给出了针对粒子群速度向量的三元离散化方法。实验表明,该优化方法得到的SOC结构参数配置,在以功耗和执行时间组成的二维目标空间中有良好的分散性和非支配性,并且该方法大大缩短了SOC结构参数的搜索时间。

关键词: 适配, 粒子群, 多目标优化, 片上系统, 参数设计