|国家预印本平台
首页|改进粒子群算法在RSoC中的软硬件划分

改进粒子群算法在RSoC中的软硬件划分

pplication of Improved Particle Swarm Optimization in Hardware/software partition of the reconfigurable system-on-chip

中文摘要英文摘要

在计算机系统结构研究中,可重构片上系统得以广泛应用,而软硬件划分是其关键的步骤之一,本文针对可重构片上系统的特点,将粒子群算法和模拟退火算法结合构造出模拟退火粒子群算法。提出一种对粒子群算法中全局最优解进行模拟退火的混合算法,避免粒子群算法陷入局部最优的危险。与基本粒子群算法和遗传算法进行比较,该混合算法具有更优化的性能,并说明了改算法对解决软硬件划分问题的有效性。

In the Computer Architecture, the reconfigurable system-on-chip was wild used.Hardware/software partitioning is an important procedure in it. According to the particular of the reconfigurable system-on-chip, in this paper, construct Simulated Annealing particle swarm algorithm by merging particle swarm algorithm with simulated annealing algorithm, which avoid trapping to the local optimum in the PSO. Based on the implementation of hybrid optimal algorithm and by comparison with the result generated by basic particle swarm algorithm and Genetic Algorithms, particle swarm embedded simulated annealing algorithm has better performance, and more effective in the problem of Hardware/software partitioning.

李仁发、吴强、刘召煜、杜家宜

计算技术、计算机技术

计算机系统结构粒子群优化模拟退火算法混合算法软硬件划分

omputer ArchitectureParticle Swarm algorithmSimulated Annealing algorithmHybrid optimal algorithmHardware/Software partitioning

李仁发,吴强,刘召煜,杜家宜.改进粒子群算法在RSoC中的软硬件划分[EB/OL].(2010-05-19)[2025-08-11].http://www.paper.edu.cn/releasepaper/content/201005-426.点此复制

评论