Academic Journal

Multi-Head Attention DQN and Dynamic Priority for Path Planning of Unmanned Aerial Vehicles Oriented to Penetration.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Multi-Head Attention DQN and Dynamic Priority for Path Planning of Unmanned Aerial Vehicles Oriented to Penetration.
Συγγραφείς: Cheng, Liuyu, Shang, Wei
Πηγή: Electronics (2079-9292); Jan2026, Vol. 15 Issue 1, p167, 35p
Θεματικοί όροι: Robotic path planning, Reinforcement learning, Decision making, Drone aircraft
Περίληψη: Unmanned aerial vehicle (UAV) penetration missions in hostile environments face significant challenges due to dense threat coverage, dynamic defense systems, and the need for real-time decision-making under uncertainty. Traditional path planning methods suffer from computational intractability in high-dimensional spaces, while existing deep reinforcement learning approaches lack efficient feature extraction and sample utilization mechanisms for threat-dense scenarios. To address these limitations, this paper presents an enhanced Deep Q-Network (DQN) framework integrating multi-head attention mechanisms with dynamic priority experience replay for autonomous UAV path planning. The proposed architecture employs four specialized attention heads operating in parallel to extract proximity, danger, alignment, and threat density features, enabling selective focus on critical environmental aspects. A dynamic priority mechanism adaptively adjusts sampling strategies during training, prioritizing informative experiences in early exploration while maintaining balanced learning in later stages. Experimental results demonstrate that the proposed method achieves 94.3% mission success rate in complex penetration scenarios, representing 7.1–17.5% improvement over state-of-the-art baselines with 2.2× faster convergence. The approach shows superior robustness in high-threat environments and meets real-time operational requirements with 18.3 ms inference latency, demonstrating its practical viability for autonomous UAV penetration missions. [ABSTRACT FROM AUTHOR]
Copyright of Electronics (2079-9292) is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Βάση Δεδομένων: Complementary Index
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:edb&genre=article&issn=20799292&ISBN=&volume=15&issue=1&date=20260101&spage=167&pages=167-201&title=Electronics (2079-9292)&atitle=Multi-Head%20Attention%20DQN%20and%20Dynamic%20Priority%20for%20Path%20Planning%20of%20Unmanned%20Aerial%20Vehicles%20Oriented%20to%20Penetration.&aulast=Cheng%2C%20Liuyu&id=DOI:10.3390/electronics15010167
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: edb
DbLabel: Complementary Index
An: 190824555
RelevancyScore: 1041
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 1041.06604003906
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Multi-Head Attention DQN and Dynamic Priority for Path Planning of Unmanned Aerial Vehicles Oriented to Penetration.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Cheng%2C+Liuyu%22">Cheng, Liuyu</searchLink><br /><searchLink fieldCode="AR" term="%22Shang%2C+Wei%22">Shang, Wei</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: Electronics (2079-9292); Jan2026, Vol. 15 Issue 1, p167, 35p
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Robotic+path+planning%22">Robotic path planning</searchLink><br /><searchLink fieldCode="DE" term="%22Reinforcement+learning%22">Reinforcement learning</searchLink><br /><searchLink fieldCode="DE" term="%22Decision+making%22">Decision making</searchLink><br /><searchLink fieldCode="DE" term="%22Drone+aircraft%22">Drone aircraft</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Unmanned aerial vehicle (UAV) penetration missions in hostile environments face significant challenges due to dense threat coverage, dynamic defense systems, and the need for real-time decision-making under uncertainty. Traditional path planning methods suffer from computational intractability in high-dimensional spaces, while existing deep reinforcement learning approaches lack efficient feature extraction and sample utilization mechanisms for threat-dense scenarios. To address these limitations, this paper presents an enhanced Deep Q-Network (DQN) framework integrating multi-head attention mechanisms with dynamic priority experience replay for autonomous UAV path planning. The proposed architecture employs four specialized attention heads operating in parallel to extract proximity, danger, alignment, and threat density features, enabling selective focus on critical environmental aspects. A dynamic priority mechanism adaptively adjusts sampling strategies during training, prioritizing informative experiences in early exploration while maintaining balanced learning in later stages. Experimental results demonstrate that the proposed method achieves 94.3% mission success rate in complex penetration scenarios, representing 7.1–17.5% improvement over state-of-the-art baselines with 2.2× faster convergence. The approach shows superior robustness in high-threat environments and meets real-time operational requirements with 18.3 ms inference latency, demonstrating its practical viability for autonomous UAV penetration missions. [ABSTRACT FROM AUTHOR]
– Name: Abstract
  Label:
  Group: Ab
  Data: <i>Copyright of Electronics (2079-9292) is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edb&AN=190824555
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/electronics15010167
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 35
        StartPage: 167
    Subjects:
      – SubjectFull: Robotic path planning
        Type: general
      – SubjectFull: Reinforcement learning
        Type: general
      – SubjectFull: Decision making
        Type: general
      – SubjectFull: Drone aircraft
        Type: general
    Titles:
      – TitleFull: Multi-Head Attention DQN and Dynamic Priority for Path Planning of Unmanned Aerial Vehicles Oriented to Penetration.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Cheng, Liuyu
      – PersonEntity:
          Name:
            NameFull: Shang, Wei
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Text: Jan2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 20799292
          Numbering:
            – Type: volume
              Value: 15
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Electronics (2079-9292)
              Type: main
ResultId 1