摘要
海上风电场的高效运行依赖于无线传感器网络提供的监测数据。通过对现有研究中传感器节点部署与通信进行调查,指出了当前海上风电场景下无线传感器节点部署通信时能效优化研究不充分的问题。针对海上风电机组无线传感器网络的能效优化问题,设计了一种基于多智能体深度确定性策略梯度算法的优化方案。考虑了节点能量有限和特定的海上通信环境特点,通过多智能体协同优化节点的感知与通信策略,有效减少能耗并提升网络覆盖率和数据传输效率。结合自适应噪声策略、优先经验回放机制以及合理的奖励函数设计,进一步提高了算法的学习效率与能效表现。实验结果表明,所提算法相比DDPG基准算法提升了约26%的节点能效,训练速度较DDQN、SAC算法加快了33%和48%。
The efficient operation of offshore wind farms relies on monitoring data from wireless sensor networks.This study identified the lack of effective energy efficiency optimization for node deployment and communication in offshore wind farm scenarios.To address this,this paper proposed an optimization scheme using a multi-agent deep deterministic policy gradient algorithm.Considering limited node energy and offshore communication challenges,the scheme optimized sensing and communication strategies through multi-agent collaboration,reducing energy consumption and improving network coverage and data transmission.Adaptive noise,prioritized experience replay,and a tailored reward function further enhanced learning efficiency and energy performance.Experiments show that the proposed scheme increases node energy efficiency by 26%over DDPG and achieves training speeds 33%and 48%faster than DDQN and SAC algorithms.
作者
贾林朋
王霄
何志琴
吴钦木
尹曜华
Jia Linpeng;Wang Xiao;He Zhiqin;Wu Qinmu;Yin Yaohua(College of Electrical Engineering,Guizhou University,Guiyang 550025,China;Guiyang Engineering Corporation Limited,POWERCHINA,Guiyang 550081,China)
出处
《计算机应用研究》
北大核心
2025年第8期2490-2496,共7页
Application Research of Computers
基金
国家自然科学基金资助项目(61861007)
贵州省科技计划资助项目(黔科合基础-ZK[2021]一般303)
贵州省科技支撑计划资助项目(黔科合支撑[2022]一般264,黔科合支撑[2023]一般096,黔科合支撑[2023]一般412,黔科合支撑[2023]一般409)
中国电建集团贵阳勘测设计研究院有限公司资助项目(YJ2022-12)
中国电力建设股份有限公司科技资助项目(DJ-ZDXM-2022-44)。
关键词
海上风电
无线传感网络
能效优化
多智能体深度强化学习
自适应噪声策略
offshore wind farms
wireless sensor network(WSN)
energy efficiency optimization
multi-agent deep reinforcement learning
adaptive noise strategy
