本文選題：內(nèi)存計(jì)算　切入點(diǎn)：面向?qū)ο箅娋W(wǎng)建模　出處：《電力系統(tǒng)自動(dòng)化》2017年11期

【摘要】：首先介紹內(nèi)存計(jì)算技術(shù)以及其在大數(shù)據(jù)處理中的應(yīng)用。針對(duì)在電力系統(tǒng)分析中計(jì)算性能的磁盤及網(wǎng)絡(luò)數(shù)據(jù)I/O瓶頸問題,提出一種基于內(nèi)存計(jì)算技術(shù)的電網(wǎng)分析內(nèi)存計(jì)算架構(gòu),并討論了內(nèi)存計(jì)算在快速動(dòng)態(tài)安全評(píng)估(DSA)中的典型應(yīng)用場景。以國調(diào)現(xiàn)有在線DSA電網(wǎng)模型(4萬節(jié)點(diǎn))為案例,給出了內(nèi)存計(jì)算在大型電網(wǎng)分析中應(yīng)用的相應(yīng)性能測試結(jié)果。
[Abstract]:This paper first introduces the memory computing technology and its application in big data processing.Aiming at the I / O bottleneck of disk and network data in power system analysis, a memory computing architecture for power network analysis based on memory computing technology is proposed.The typical application scenarios of memory computing in DSAs are also discussed.Taking the existing on-line DSA power network model (40000 nodes) as an example, the corresponding performance test results of the application of memory computing in large-scale power network analysis are given.
【作者單位】： 南瑞集團(tuán)公司(國網(wǎng)電力科學(xué)研究院)(北京);華南理工大學(xué)電力學(xué)院;
【基金】：國家電網(wǎng)公司“千人計(jì)劃”專項(xiàng)資助項(xiàng)目(5206001600A3)~~
【分類號(hào)】：TM711
，

本文編號(hào)：1730484