首页|基于强化学习的资源受限传感器信号智能采集研究

基于强化学习的资源受限传感器信号智能采集研究

扫码查看
针对资源受限传感器,在自主条件下的信号样本采集问题,设计了资源受限传感器的信号采集模型,提出了基于强化学习的信道采集选择算法,从而提高了信号样本的有效采集效率.文章比较了基于Q-Learning、DQN、DDPG等强化学习算法在信道选择应用的学习效果.通过仿真试验验证,在设置适当超参数的条件下,传感器信号样本的自主采集效率可提升至接近理论极限的95%以上,证明设计的智能采集模型可有效提升传感器自主采集的智能水平.
Resource-Constrained Sensor Signal Collecting Method Based on Reinforcement Learning Algorithm
For the signal collecting efficiency problem of the resource-constrained sensors under autonomous conditions,the sig-nal collecting model of the resource-constrained sensors was proposed.Through the collecting signal channel selection algorithm based on reinforcement learning is proposed to improve the effective collection efficiency of signal samples.The learning effects of Q-Learn-ing,DQN and DDPG method were also compared in this paper.The experiment results show that the proposed method can provide high efficiency of signal samples collecting.With suitable super parameters,the collecting efficiency can be promoted to the 95%val-ue of the theoretical limits,which proves that the intelligent collection model can effectively improve the intelligent level of sensor au-tonomous collection.

Q-LearningDQNreinforcement learningsensorsignal collecting

叶李

展开 >

中国西南电子技术研究所,成都 610036

Q-Learning DQN 强化学习 传感器 信号采集

国家自然科学基金面上项目

62072077

2024

自动化与仪器仪表
重庆工业自动化仪表研究所,重庆市自动化与仪器仪表学会

自动化与仪器仪表

CSTPCD
影响因子:0.327
ISSN:1001-9227
年,卷(期):2024.(1)
  • 7