航空学报 > 2025, Vol. 46 Issue (12): 231448-231448   doi: 10.7527/S1000-6893.2024.31448

固体力学与飞行器总体设计

基于DQN的无人机主动捕捉拦阻回收方法

王一峰1,2, 彭一明1,2(), 李龙1,2, 魏小辉1,2, 聂宏1,2   

  1. 1.南京航空航天大学,航空航天结构力学及控制全国重点实验室,南京 210016
    2.南京航空航天大学,飞行器先进设计技术国防重点学科实验室,南京 210016
  • 收稿日期:2024-10-28 修回日期:2024-11-11 接受日期:2024-12-03 出版日期:2024-12-11 发布日期:2024-12-10
  • 通讯作者: 彭一明 E-mail:yimingpeng@nuaa.edu.cn
  • 基金资助:
    江苏省自然科学基金(BK20220910);国家自然科学基金(52202441);国家自然科学基金(52275114);国防卓越青年科学基金(2018-JCJQ-ZQ-053);航空科学基金(20240013052003)

DQN-based active arrest and recovery technique for UAVs

Yifeng WANG1,2, Yiming PENG1,2(), Long LI1,2, Xiaohui WEI1,2, Hong NIE1,2   

  1. 1.State Key Laboratory of Mechanics and Control for Aerospace Structures,Nanjing University of Aeronautica and Astronautics,Nanjing 210016,China
    2.Key Laboratory of Fundamental Science for National Defense-Advanced Design Technology of Flight Vehicle,Nanjing University of Aeronautica and Astronautics,Nanjing 210016,China
  • Received:2024-10-28 Revised:2024-11-11 Accepted:2024-12-03 Online:2024-12-11 Published:2024-12-10
  • Contact: Yiming PENG E-mail:yimingpeng@nuaa.edu.cn
  • Supported by:
    Natural Science Foundation of Jiangsu Province(BK20220910);National Natural Science Foundation of China(52202441);National Defense Outstanding Youth Science Foundation(2018-JCJQ-ZQ-053);Aeronautical Science Foundation of China(20240013052003)

摘要:

使用拦阻索回收无人机时,钩索成功率是评估其回收安全性与可靠性的关键指标之一。针对无人机回收区域受限的情况下如何提高钩索成功率这一问题,提出了一种无人机主动捕捉拦阻回收方法,通过在车载移动平台的拦阻系统在人工智能计算结果的引导下主动移动至最佳钩索位置来提高无人机的钩索成功率。首先,建立无人机着陆回收拦阻动力学模型来计算钩索失效边界,使用支持向量机(SVM)方法对动力学仿真结果进行识别,并生成无人机钩索分析代理模型。然后,将移索过程简化成马尔可夫决策过程,使用移索装置模型作为训练环境,六自由度无人机着陆下滑模型生成数据集,钩索分析代理模型构成奖励函数,采用深度Q网络(DQN)训练得到能够实时计算并引导拦阻装置向最佳钩索位置主动调整的移索策略。仿真结果显示,在回收空间受限的情况下,与传统被动式拦阻回收方法相比,使用主动捕捉拦阻方法钩索成功率提高了29%。该方法有效提高了无人机着陆回收的安全性和可靠性,为智能化回收技术的发展提供了新的理论支持和实践方案。

关键词: 固定翼无人机, 着陆回收, 钩索成功率, DQN, 人工智能

Abstract:

The success rate of hooking the cable is one of the key indicators of safety and reliability when recovering a UAV using arresting cables. To address the challenge of improving the success rate with restricted UAV recovery areas provided by vehicle-mounted mobile platform, this paper proposes an active cable engagement method for UAV arresting and recovery. The method enhances the success rate by having the arresting system actively move to the optimal cable engagement position under the guidance of AI computation results. First, a UAV landing and recovery dynamics model is established to calculate the cable engagement failure boundary. The Support Vector Machine (SVM) method is applied to identify the dynamics simulation results and generate a proxy model for UAV cable engagement analysis. Then, the cable-moving process is simplified into a Markov decision process. Using the cable-moving device model as the training environment, a six-degree-of-freedom UAV landing and descent model is employed to generate the dataset, and the cable engagement analysis proxy model is used as the reward function. A Deep Q-Network (DQN) is applied to train a cable-moving strategy that dynamically computes and guides the arresting system to adjust to the optimal engagement position in real time. Simulation results show that with limited recovery space, the proposed method improves the success rate by 29% compared to the traditional passive recovery method. This approach significantly enhances the safety and reliability of UAV landing and recovery, providing new theoretical support and practical solutions for the development of intelligent recovery technology.

Key words: fixed-wing UAV, landing recovery, success rate of hooking cable, deep Q-network (DQN), artificial intelligence (AI)

中图分类号: