摘要
在云边协同环境中,资源状态动态变化,固定分配策略易导致资源浪费。文章提出基于深度强化学习的动态任务调度方法,通过智能管理任务队列和优先级评估,使任务高效有序执行。其中,调度模型能够自动学习与优化策略,提高资源利用率。实验结果表明,该方法使资源利用率稳步提升,最终接近100%,验证了其有效性及在云边协同调度中的应用潜力。
In cloud-edge collaborative environments,resource states change dynamically,making fixed allocation strategies prone to resource waste.This article proposes a dynamic task scheduling method based on deep reinforcement learning.By intelligently managing task queues and evaluating priorities,it enables efficient and orderly task execution.The scheduling model can automatically learn and optimize strategies,thereby improving resource utilization.Experimental results show that this method steadily increases resource utilization,eventually approaching 100%,verifying its effectiveness and application potential in cloud-edge collaborative scheduling.
作者
林志达
林克全
吴文鹏
纪伟
王可
LIN Zhida;LIN Kequan;WU Wenpeng;JI Wei;WANG Ke(China Southern Power Grid Co.,Ltd.,Guangzhou 510450,China;China Southern Power Grid Digital Grid Research Institute Co.,Ltd.,Guangzhou 525299,China;Guangdong Provincial Key Laboratory of Digital Power Grid Technology,Guangzhou 525299,China)
基金
中国南方电网有限责任公司科技项目(ZBKJXM20232454)。
关键词
深度强化学习
云边协同
动态任务调度
云边协同动态任务
任务调度
deep reinforcement learning
cloud-edge collaboration
dynamic task scheduling
cloud-edge collaborative dynamic task
task scheduling
