南京邮电大学学报（自然科学版）2025，Vol.45Issue(3)：87-98,12.DOI:10.14132/j.cnki.1673-5439.2025.03.010
基于云边协同的智能电网多无人机辅助计算迁移
Multi-UAV assisted computation offloading in smart grid based on cloud-edge collaboration
摘要
Abstract
The application of edge computing in smart grid can partially meet the real-time requirements of power industry.However,as the grid is growing and the density of power terminal devices is rising,the statically deployed edge nodes can hardly cover all the scattered terminal devices,leading to extra-low computation offloading performance and efficiency.A multi-unmanned aerial vehicle(UAV)-assisted computational offloading algorithm based on double delay deep deterministic policy gradient is proposed and named as collaborative cloud-edge computing offloading based on TD3(CeCO-TD3).This algorithm constructs a multi-objective optimization function that incorporates UAV flight angle,distance,offload-ing service selection and task offloading ratio and aims at minimizing the system's computational delay and energy consumption.The optimization problem is solved by using a deep reinforcement learning algo-rithm.A cloud-side collaboration framework and a cloud-based policy experience pool with priorities are also introduced to further guarantee the quality of computational offloading services for multiple UAVs.Experimental results show that the proposed algorithm outperforms the traditional optimization algorithms in terms of scaling down the task transmission delay and reducing the computational energy consumption.关键词
边缘计算/智能电网/深度强化学习/计算迁移Key words
edge computing/smart grid/deep reinforcement learning/computational offloading分类
计算机与自动化引用本文复制引用
刘尚东,邱华杰,黄昕,沙宇晨,季一木..基于云边协同的智能电网多无人机辅助计算迁移[J].南京邮电大学学报（自然科学版）,2025,45(3):87-98,12.基金项目
国家重点研发计划(2023YFB2904000,2023YFB2904004)、江苏省重点发展规划项目(BE2023004-2)、江苏省自然科学及高校自然科学重大项目(20KJA520001)、中国电信江苏分公司2023年科技项目(JSSGS2301022EGN00)和未来网络科学研究基金(FNSRFP-2021-YB-15)资助项目 （2023YFB2904000,2023YFB2904004）
