首页|期刊导航|电气传动|边缘计算架构下考虑电力业务特性的缓存配置方法

边缘计算架构下考虑电力业务特性的缓存配置方法OA

Cache Configuration Method Considering Power Service Characteristics in Edge Computing Architecture

中文摘要英文摘要

随着新型电力系统向高比例可再生能源接入与多元互动化方向演进,边缘计算作为支撑电力系统智能化升级的核心技术,正面临着设备规模化接入与电力业务的复杂化的严峻挑战.针对传统静态资源配置模式难以适配电力业务动态性、异构性特征的问题,以缓存资源对电力业务执行的弹性作用为切入点,提出了一种考虑电力业务特性的缓存资源配置方法.以业务时延、资源利用率最优为配置目标,系统性地分析电力业务特性对资源配置的影响机制,构建了业务特性驱动的约束条件体系,并基于权重分析法对该资源配置问题进行求解.通过建立并对比不同的配置方案下的资源和业务评价指标,充分证明了提出的方法的可行性和优越性.

As the new power system evolves towards a high proportion of renewable energy access and multiple interactions,edge computing,as the core technology to support the intelligent upgrade of power system,is facing severe challenges of large-scale equipment access and complexity of power business.Aiming at the problem that the traditional static resource allocation model is difficult to adapt to the dynamic and heterogeneous characteristics of power services,the elastic effect of cache resources on the execution of power services was taken as the entry point,and a cache resource allocation method was proposed considering the characteristics of power services.Aiming at the optimization of service delay and resource utilization,the influence mechanism of power service characteristics on resource allocation was analyzed systematically,a constraint condition system driven by service characteristics was constructed,and the resource allocation problem was solved based on weight analysis method.By establishing and comparing the resource and business evaluation indicators under different configuration schemes,the feasibility and superiority of the proposed method were fully proved.

陈柏旭;李晓华;蔡泽祥;胡明;梁旭冉;阳斌

华南理工大学电力学院,广东 广州 510030华南理工大学电力学院,广东 广州 510030华南理工大学电力学院,广东 广州 510030华南理工大学电力学院,广东 广州 510030华南理工大学电力学院,广东 广州 510030华南理工大学电力学院,广东 广州 510030

信息技术与安全科学

边缘计算电力业务特性弹性需求缓存配置

edge computingpower service characteristicselastic demandcache configuration

《电气传动》 2026 (5)

61-69,9

广东省基础与应用基础研究基金(2024B1515250001)广东省自然科学基金(2025A1515010058)国家自然科学基金(52477098)

10.19457/j.1001-2095.dqcd26470

评论