基于计算特征的分布式空三计算任务调度研究

该文从倾斜摄影测量的现状出发,整理主流空三算法的计算流程,通过分析其算法特性,结合计算过程中使用的硬件资源,为提升计算过程中的任务调度效率,设计基于计算特征的分布式空三计算任务调度规则。并针对设计的调度规则开展一系列对比实验。实验结果表明,在梯度分布的计算节点上,使用基于计算特征的分布式空三计算任务调度规则的调度方式与不使用调度规则的方式相比,由于高性能主机在任务派送的优先级较高,因而任务包的周转度更高,使用调度规则的分布式集群在计算时长的表现上更好,在空三计算效率的提升方向及提升手段上具有一定的参考意义。

Starting with the present situation of oblique photo grammetry,this paper sorts out the calculation flow of the mainstream spatial three algorithms,through the analysis of its algorithm characteristics,based on the hardware resources used in the computing process,in order to improve the efficiency of task scheduling in the computing process,a distributed spatial three computing task scheduling rule based on computing features is designed.A series of comparative experiments are carried out according to the designed scheduling rules.The experimental results show that on the computing nodes with gradient distribution,the scheduling method using distributed spatial three computing task scheduling rules based on computing features is compared with that without scheduling rules.Since the high-performance host has a higher priority in task delivery,hence the higher turnover of the task package.The distributed cluster using scheduling rules performs better in computing time,and has a certain reference significance in the direction and means of improving the efficiency of empty three computing.