网格资源提前预留中用户资源需求量预测模型

A prediction model to quantify user′s demand in grid resource reservation

针对目前对网格资源提前预留的研究主要集中在资源预留算法上,几乎忽略了如何估算用户资源需求量问题,提出一种适用于网格资源预留的用户资源需求量预测模型.基于对网格资源提前预留请求的定义,结合GM(1,1)灰色预测方法,建立了用户资源需求量预测模型,并提出相应的改进方法,提高了预测精度.模拟实验证明了模型的有效性,能获得较好的预测效果.

According to the problem that the literatures′ main focus on grid resource reservation is reservation algorithms,while how to quantify the resource requirements of users is almost ignored.Based on the definition of advance resource reservation inquiry of Grids,and by means of GM(1,1) prediction method,a prediction model for quantifying the resource requirements of users was presented.Furthermore,in order to improve prediction accuracy,a modified prediction strategy was also presented.At last,simulation results showed the validity and effectiveness of our model,as well as the better prediction effectiveness.