青海师范大学学报(自然科学版）

在“东数西算”与“双碳”战略双重背景下,科学预测全国算力需求趋势、精准评估区域绿色算力发展潜力,对优化国家算力资源配置、推动区域差异化发展具有重要的现实意义.本研究基于2018—2024年全国算力数据构建ARIMA模型预测需求趋势,同时通过主成分分析法对西部九省(区、市)绿色算力发展进行评估,结果显示青海省在绿色资源禀赋方面位居西部首位,呈现“高禀赋、弱设施”的结构特征;进一步运用动态主成分分析法对青海、内蒙古、贵州三省(区)的演进轨迹进行追踪,研究发现青海省自2022年起产业动能显著加速,正处于从“资源潜力”向“产业竞争力”转换的关键阶段.据此,本研究提出青海省应立足绿电资源优势,走“以绿电为核心、以零碳为标志”的差异化发展路径,为国家算力资源优化布局与区域绿色低碳协同发展提供决策参考.

Under the dual background of the "Eastern Data and Western Computing" project and the "Double Carbon" strategy, scientifically predicting the trend of national computing power demand and accurately evaluating the development potential of regional green computing power are of great practical significance for optimizing the allocation of national computing power resources and promoting regional differentiated development.Based on national computing power data from 2018 to 2024,this study constructed an ARIMA model to predict the demand trend,and simultaneously evaluated the development of green computing power in nine western provinces through principal components analysis(PCA).The results show that Qinghai Province ranks first in western China in terms of green resource endowment, presenting a structural characteristic of "high endowment and weak infrastructure".By further using dynamic principal components analysis(DPCA) to track the evolution trajectories of Qinghai,Inner Mongolia and Guizhou, it is found that Qinghai's industrial kinetic energy has accelerated significantly since 2022, and it is in a critical stage of transformation from "resource potential" to "industrial competitiveness".Accordingly, this study proposes that Qinghai should base on the advantages of green power resources and take a differentiated development path "with green power as the core and zero carbon as the symbol", so as to provide decision-making references for the optimal layout of national computing power resources and the coordinated development of regional green and low-carbon economy.