【摘要】 文章提出了一种K均值聚类和模拟退火融合的软硬件划分算法。算法首先将有相似属性的任务节点通过K均值聚类算法组成一个大的任务节点,而后使用模拟退火算法划分由大的任务节点组成的系统。通过对比经典的模拟退火软硬件划分技术以及实验结果的验证表明,使用K均值聚类和模拟退火融合的软硬件划分算法使有着较多任务节点的复杂系统的软硬件划分快速收敛到合适的值。更多还原

【Abstract】 This paper proposes a hardware/software partitioning algorithm of embedded system based on K-means clustering and simulated annealing.First,this algorithm assembles task vertex with similar attribute to form a bigger new task vertex based on K-means clustering,and then partitions hardware/software based on simulated annealing algorithm.The experiments by contrasting to the classic simulated annealing algorithm have showed that using this algorithm can accelerate convergence of complex embedded system with more task vertex.更多还原