摘要： 在分布式图像绘制中,为提高图像绘制速度,达到图像的实时绘制,缩短任务调度长度,提出了基于最早完成时间(EFT)启发的遗传算法,染色体编码采用问题属性作为基因。实验表明它对于解决异构网络平台下分布式任务调度具有很好的收敛速度,最佳的调度长度和调度方案。应用于分布式图像绘制可以取得良好的实时性和较佳的图像质量。

关键词: 遗传算法, 异构分布式计算机, 任务调度, 分布式图像绘制, 最早完成时间

Abstract: In distributed image rendering, for improving image rendering speed, achieving image real-time rendering, shortening the task schedule length, the paper proposes genetic algorithm based on the earliest finish time(EFT) heuristic, it adopts attribute of problem as chromosome encoding. Experiment result indicates that it has fine convergence speed, good scheduling length, and fine scheduling scheme for distributed task scheduling in solution to heterogeneous network platform. It has good real-time for distributed image rendering and fine image quality.

中图分类号: 

• TP391.9