摘要
针对云计算中资源有效分配的问题,提出一种基于改进型离散粒子群优化(IDPSO)算法的云资源分配方案.首先,将传统PSO算法中的运算进行离散化,使其能够应用于资源分配问题.然后,对传统PSO粒子位置更新公式中的惯性权重进行改进,根据当前粒子位置、局部最佳和全局最佳位置的适应度来确定这些权重系数,以此加快粒子的收敛速度.最后,将资源分配方案编码为一个二维粒子,利用IDPSO算法求解最优解.实验结果表明,该方案能够有效降低资源浪费率,具有可行性和有效性.
For the issue of the resource al locat ion in c loud comput ing, a cloud comput ing resources al location scheme based on improved discrete part icle swarm opt imizat ion (IDPSO) algori thm is proposed. Firstly, the operat ion of the t radi t ional PSO algori thm is discret ized, so that it can be appl ied to the resource allocat ion. Then, the inert ia weight in part icle posit ion update formula of tradi tional PSO is improved, which is determined by the adaptat ion of the part icle current, local best and global best position, in order to speed up the convergence speed of par t icle. Final ly, the resource al locat ion scheme is coded as a two-dimensional part icle, and the opt imal solut ion is solved by the IDPSO algori thm. Experimental results show that the proposed scheme can ef fect ively reduce the waste of resources, and it is feasible and effective.
出处
《湘潭大学自然科学学报》
北大核心
2017年第3期89-93,共5页
Natural Science Journal of Xiangtan University
基金
江西省教育厅科学技术研究项目(151574)
关键词
云计算
资源分配
改进型离散粒子群优化
二维粒子编码
cloud comput ing
resources al locat ion
improved dis c rete par t icle swarm opt imizat ion
two dimensional part icle coding