重庆邮电大学学报（自然科学版）2025，Vol.37Issue(4)：507-515,9.DOI:10.3979/j.issn.1673-825X.202410100249
电力物联网自适应任务卸载策略
Adaptive task offloading strategy for power internet of things
摘要
Abstract
To address issues such as unbalanced resource allocation,long computing latency,and excessive energy con-sumption in mobile edge computing for the power internet of things(PIoT),this paper proposes a dynamic resource alloca-tion strategy based on reinforcement learning and convolutional neural networks.By making real-time decisions to optimize task offloading,the proposed method significantly enhances computing performance and energy efficiency.Considering fac-tors such as resource availability,task characteristics,and network conditions,a dynamic resource allocation mechanism is developed,transforming the allocation problem into a dynamic optimization problem.A deep reinforcement learning-based adaptive offloading algorithm is then introduced to solve this problem.Simulation results demonstrate that the proposed algo-rithm achieves notable improvements in task completion rate,resource utilization,and system adaptability.Under highly volatile network conditions,task completion rate improves by 15%,and resource utilization increases by 10%.关键词
电力物联网/边缘计算/深度学习/神经网络/资源分配Key words
power internet of things/edge computing/deep learning/neural networks/resource allocation分类
信息技术与安全科学引用本文复制引用
贺超,汪超凡,张思睿,张一,王碧..电力物联网自适应任务卸载策略[J].重庆邮电大学学报（自然科学版）,2025,37(4):507-515,9.基金项目
重庆市万州区科研项目(wzstc20230315,wzstc20230418) （wzstc20230315,wzstc20230418）
重庆市教委科学技术研究项目(KJZD-K202201203,KJQN202301258) （KJZD-K202201203,KJQN202301258）
重庆市自然科学基金创新发展联合基金项目(CSTB2022NSCQ-LZX0055) Wanzhou District Scientific Research Project of Chongqing(wzstc20230315,wzstc20230418) （CSTB2022NSCQ-LZX0055）
Scientific and Tech-nological Research Project of Chongqing Municipal Education Commission(KJZD-K202201203,KJQN202301258) （KJZD-K202201203,KJQN202301258）
Chongqing Natural Science Foundation Innovation and Development Joint Fund Project(CSTB2022NSCQ-LZX0055) （CSTB2022NSCQ-LZX0055）
