摘要
针对深度强化学习方法在复杂环境下面对无人机博弈任务时学习效率较低的问题,提出了知识和数据联合驱动的深度强化学习模型。首先,借鉴了模仿学习的思想,将遗传算法作为启发式搜索策略,并收集专家经验知识;其次,通过深度强化学习与环境进行交互,收集在线经验数据;最后,构建了知识和数据联合驱动的深度强化学习模型,用于优化无人机博弈策略。实验结果表明,所提模型有效提升了收敛速度和学习稳定性,经过训练的智能体具有较好的自主博弈路径规划能力。
A deep reinforcement learning model driven by knowledge and data was proposed to address the low learning efficiency of deep reinforcement learning methods in complex environments for unmanned aerial vehicle(UAV)game tasks.Firstly,drawing on the idea of imitation learning,a genetic algorithm was employed as a heuristic search strategy,and expert experience knowledge was collected.Secondly,the UAV interacted with the environment through deep reinforcement learning and collected online experi-ence data.Finally,a deep reinforcement learning model driven by knowledge and data was constructed to optimize UAV game strategies.Experimental results indicated that the proposed model effectively im-proved the convergence speed and learning stability,and the trained agents demonstrated better autono-mous game path planning capabilities.
作者
薛均晓
张世文
陆亚飞
严笑然
付玮
XUE Junxiao;ZHANG Shiwen;LU Yafei;YAN Xiaoran;FU Wei(School of Cyber Science and Engineering,Zhengzhou University,Zhengzhou 450002,China;Institute of Artificial Intelligence,Zhejiang Lab,Hangzhou 311100,China;Research Center of Space Computing,Zhejiang Lab,Hangzhou 311100,China)
出处
《郑州大学学报(理学版)》
北大核心
2025年第4期8-14,共7页
Journal of Zhengzhou University:Natural Science Edition
基金
国家重点研发计划项目(2022YFC3004400)。
关键词
深度强化学习
无人机博弈
路径规划
遗传算法
deep reinforcement learning
UAV game
path planning
genetic algorithm
