摘要
在电力系统调度自动化领域,随着数据规模的不断变大,电力系统计算会变得复杂耗时,提出一种基于电力系统大规模数据集的并行计算方法。基于云平台技术,利用容器轻量级虚拟应用的特点,以电力系统短路计算为例,采用将短路计算节点矩阵并行分解算法,将大规模的电力系统数据集,分解成多个可独立计算的数据单元,根据节点规模动态创建和释放短路计算容器完成短路计算并返回结果。通过Cloudsim对不同节点规模的短路计算进行云平台仿真,对比了串行算法、虚拟机并行算法和容器并行算法的计算时间,结果表明采用该方法可以大幅度提升短路计算并行化的效率。
With the increasing data size in the field of power system,power system calculation will become complex and time-consuming.This paper proposes a parallel computing method based on large-scale data sets of power system.For the short circuit calculation in cloud platform of power system analysis,this paper proposes a cloud platform short-circuit calculation parallelization methods based on Docker to improve the efficiency and speed of the cloud computing.This method takes advantage of the lightweight virtual application characteristics of Docker.The node matrix parallel decomposition algorithm for short circuit calculation is used.The container could dynamically create and release to complete short circuit calculation according to the node size.Cloud platform simulation of short-circuit computing of different node sizes is carried out through Cloudsim.The computing time of serial algorithm,virtual machine parallel algorithm and container parallel algorithm is compared.The results show that the method in this paper can greatly improve the efficiency of short-circuit computing parallelization.
出处
《工业控制计算机》
2021年第2期86-87,90,共3页
Industrial Control Computer
关键词
并行计算
云平台
容器化
电力自动化
parallel computing
cloud platform
Docker
power automation
