考虑计算任务端到端完成时延的算力-电力协调优化运行方法
Optimal Operation Method for Computing Power-Electric Power Coordination Considering End-to-End Completion Latency of Computing Tasks
浏览论文内容
中文总结 AI 辅助
本文提出一种考虑计算任务端到端完成时延的算力-电力协调优化运行方法，通过建立全流程时延模型和任务级时空调度机制，联合优化任务转发时间、路由路径和目的节点，以最小化供电成本与任务总时延，实现计算负荷与可再生能源的时空匹配。
中文摘要 AI 辅助
随着计算需求的快速增长和可再生能源的大规模并网，如何实现计算网络与电力网络的协调优化运行已成为亟待解决的重要问题。然而，现有研究主要关注计算负荷时空迁移对电力系统运行的影响，而计算网络内部的任务处理流程在很大程度上被忽视。针对这一问题，本文提出了一种考虑计算任务端到端完成时延的算力-电力协调优化运行方法。首先，建立了覆盖计算任务“传输-缓冲-计算”全过程的全流程时延模型，该模型统一刻画了转发等待、网络传输、排队和计算处理等环节。其次，开发了一种任务级时空调度机制，以联合优化任务转发时间、路由路径和目的计算节点。然后，构建了算力-电力协调优化模型，以最小化电力网络的供电成本和计算任务的总完成时延。案例研究表明，所提方法能够充分利用计算负荷的任务级时空调度灵活性，促进计算负荷与可再生能源之间的时空匹配，从而在保障计算任务服务质量的同时降低系统供电成本。
英文摘要
With the rapid growth of computing demand and the large-scale integration of renewable energy, how to realize the coordinated optimal operation of computing networks and power networks has become an important issue to be addressed. However, existing studies mainly focus on the impacts of spatiotemporal migration of computing loads on power system operation, while the internal task processing procedures of computing networks are largely neglected. To address this issue, this paper proposes an optimal operation method for computing power-electric power coordination considering the end-to-end completion latency of computing tasks. First, a full-process latency model covering the "transmission-buffering-computation" procedure of computing tasks is established, which uniformly characterizes forwarding waiting, network transmission, queueing, and computation processing. Second, a task-level spatiotemporal scheduling mechanism is developed to jointly optimize the task forwarding time, routing path, and destination computing node. Then, a computing power-electric power coordinated optimization model is formulated to minimize the power supply cost of the power network and the total completion latency of computing tasks. Case studies demonstrate that the proposed method can fully exploit the task-level spatiotemporal scheduling flexibility of computing loads and facilitate the spatiotemporal matching between computing loads and renewable energy, thereby reducing the system power supply cost while ensuring the quality of service for computing tasks.
发表机构
- State Key Laboratory of Advanced Electromagnetic Engineering and Technology (School of Electrical and Electronic Engineering, Huazhong University of Science and Technology)（电磁场与电器技术国家重点实验室（华中科技大学电气与电子工程学院））
- Department of Electrical Engineering and Electronics, University of Liverpool（利物浦大学电气电子工程系）
- School of Electrical Engineering, Southeast University（东南大学电气工程学院）
- State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology（华中科技大学煤燃烧国家重点实验室）
- Warsaw University of Technology（华沙理工大学）
机构由 AI 辅助整理，请以论文原文为准。
