Scholarlink

The integration of Brain-Computer Interface (BCI) technology with advanced Artificial Intelligence (AI) techniques presents a transformative approach to enhancing the autonomy and human-machine interaction of Unmanned Aerial Vehicles (UAVs). Traditional UAV control systems, reliant on manual remote operation or pre-programmed paths, face limitations in dynamic, high-risk environments such as search and rescue, disaster response, and medical delivery. This study addresses these challenges by developing a novel control algorithm that decodes electroencephalography (EEG) signals in real time to enable intuitive, hands-free UAV navigation. By leveraging Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRUs) for EEG signal classification and reinforcement learning (RL) for autonomous flight optimization, the proposed system translates cognitive intent into precise UAV maneuvers. Rigorous testing in the AirSim simulation platform demonstrates significant improvements in classification accuracy (94.7%), response latency (<250 ms), and adaptability to environmental variability. While simulation results are promising, we acknowledge that future real-world deployment must address critical challenges such as environmental interference, UAV hardware constraints, and security risks to ensure robust, reliable operation. Overall, the results underscore the system’s potential for real-world applications, offering a scalable, low-latency solution that bridges human cognitive intent with autonomous UAV control.

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

Brain-Controlled UAV Navigation: A Deep Learning and Reinforcement Learning Framework for Real-Time Aerial Maneuvering

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Luma Hameed Hadi,
  • B. M. Albaker

摘要

The integration of Brain-Computer Interface (BCI) technology with advanced Artificial Intelligence (AI) techniques presents a transformative approach to enhancing the autonomy and human-machine interaction of Unmanned Aerial Vehicles (UAVs). Traditional UAV control systems, reliant on manual remote operation or pre-programmed paths, face limitations in dynamic, high-risk environments such as search and rescue, disaster response, and medical delivery. This study addresses these challenges by developing a novel control algorithm that decodes electroencephalography (EEG) signals in real time to enable intuitive, hands-free UAV navigation. By leveraging Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRUs) for EEG signal classification and reinforcement learning (RL) for autonomous flight optimization, the proposed system translates cognitive intent into precise UAV maneuvers. Rigorous testing in the AirSim simulation platform demonstrates significant improvements in classification accuracy (94.7%), response latency (<250 ms), and adaptability to environmental variability. While simulation results are promising, we acknowledge that future real-world deployment must address critical challenges such as environmental interference, UAV hardware constraints, and security risks to ensure robust, reliable operation. Overall, the results underscore the system’s potential for real-world applications, offering a scalable, low-latency solution that bridges human cognitive intent with autonomous UAV control.