Scholarlink

<p>This paper presents a scalable and robust framework for unmanned aerial vehicle to unmanned aerial vehicle (UAV-UAV) attacker-defender interception strategy by integrating sliding mode backstepping control along with unsupervised machine learning in the presence of disturbances in the defender dynamics. The proposed approach combines real time clustering (K-mean clustering) to efficiently organize adversary UAVs into tactical clusters and implement a Lyapunov-stable sliding mode backstepping controller to guarantee accurate and resilient interception amidst real-world uncertainty. Furthermore, PID control was used to perform the interception of the attacker centroids in the presence of disturbances. The interception strategy is well validated by two case studies. First, a single UAV defender engages three UAV attackers, successfully intercepting their centroid through early target interception and minimal convergence error. Secondly, the scenario involves three defenders intercepting nine attacking UAVs by targeting three centroids derived by clustering, achieving complete interception with minimal error. This hierarchical clustering-control architecture minimizes computational complexity from numerous attacking UAVs to their centroids, ensuring chattering-free and minimal overshoot control inputs with guaranteed stability using Lyapunov analysis. Simulation results validate the framework’s capability for real-time deployment. The proposed approach demonstrates effective interception strategies in terms of speed and resource efficiency, while ensuring robustness in dynamic situations. While the PID control interception mission fails showing clearly the robustness and superiority of the proposed control interception strategy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Sliding mode based backstepping control for UAV-UAV interception strategy with K-mean clustering

  • 期刊论文

  • 发表日期：

    2025年9月20日

查看全文
  • Abdulrazaq Nafiu Abubakar,
  • Ali Nasir,
  • Abdul-Wahid A. Saif

摘要

This paper presents a scalable and robust framework for unmanned aerial vehicle to unmanned aerial vehicle (UAV-UAV) attacker-defender interception strategy by integrating sliding mode backstepping control along with unsupervised machine learning in the presence of disturbances in the defender dynamics. The proposed approach combines real time clustering (K-mean clustering) to efficiently organize adversary UAVs into tactical clusters and implement a Lyapunov-stable sliding mode backstepping controller to guarantee accurate and resilient interception amidst real-world uncertainty. Furthermore, PID control was used to perform the interception of the attacker centroids in the presence of disturbances. The interception strategy is well validated by two case studies. First, a single UAV defender engages three UAV attackers, successfully intercepting their centroid through early target interception and minimal convergence error. Secondly, the scenario involves three defenders intercepting nine attacking UAVs by targeting three centroids derived by clustering, achieving complete interception with minimal error. This hierarchical clustering-control architecture minimizes computational complexity from numerous attacking UAVs to their centroids, ensuring chattering-free and minimal overshoot control inputs with guaranteed stability using Lyapunov analysis. Simulation results validate the framework’s capability for real-time deployment. The proposed approach demonstrates effective interception strategies in terms of speed and resource efficiency, while ensuring robustness in dynamic situations. While the PID control interception mission fails showing clearly the robustness and superiority of the proposed control interception strategy.