location: Current position: Home >> Scientific Research >> Paper Publications

一种基于GPU加速的细粒度并行粒子群算法

Hits:

Indexed by:期刊论文

Date of Publication:2006-12-30

Journal:哈尔滨工业大学学报

Included Journals:EI、PKU、ISTIC、CSCD、Scopus

Volume:38

Issue:12

Page Number:2162-2166

ISSN No.:0367-6234

Key Words:粒子群优化算法;并行处理;GPU;细粒度

Abstract:提出了一种基于GPU加速的细粒度并行粒子群算法,将并行PSO求解过程转化为GPU纹理渲染过程,使PSO算法在GPU中加速执行,在取得了较好的优化效果的同时,增大了细粒度并行的粒子规模,提高了算法的运算速度,并为普通用户的并行PSO工作提供了一种可行的方法.

Pre One:一种快速灰度图像彩色化算法

Next One:The maximum variance between clusters method of image segmentation based on particle swarm optimization