摘要
为改善蚁群算法对大规模旅行商问题的求解性能,提出一种基于图形处理器(GPU)加速的细粒度并行蚁群算法.将并行蚁群算法求解过程转化为统一计算设备架构的线程块并行执行过程,使得蚁群算法在GPU中加速执行.实验结果表明,该算法能提高全局搜索能力,增大细粒度并行蚁群算法的蚂蚁规模,从而提高了算法的运算速度.
An algorithm of fine-grained parallel ant colony optimization algorithm （ACO） based on graphics process unit（GPU） accelerated is proposed to improve the performance of ACO for application to large-scale TSP problems. The process of parallel ACO is convert into that of parallel compute unified device architecture （CUDA） thread blocks, which makes PACO speed up. The experimental results show that the algorithm improves the ability of global search, increases the ant population in the PACO, speeds up its running and provides ordinary user with a feasible PACO solution.
出处
《控制与决策》
EI
CSCD
北大核心
2009年第8期1132-1136,共5页
Control and Decision
基金
国家自然科学基金项目(70571009
70671014)
国家杰出青年基金项目(70725004)
高等学校博士点基金项目(20060141013)
辽宁省高等学校优秀人才支持计划项目([2006]124)
关键词
蚁群算法
并行处理
图形处理器
细粒度
Ant colony optimization algorithm
Parallel process
Graphics process unit
Fine-grained
