This study proposes a framework that integrates a fine-tuned Large Language Model (LLM) with distributed Model Predictive Control (MPC) for efficient drone swarm control in dynamic environments. An LLM, trained on over 4,000 command data samples, tran...
This study proposes a framework that integrates a fine-tuned Large Language Model (LLM) with distributed Model Predictive Control (MPC) for efficient drone swarm control in dynamic environments. An LLM, trained on over 4,000 command data samples, translates natural language commands into structured
mission plans. Subsequently, the local MPC on each drone optimizes flight trajectories while performing collision avoidance and formation maintenance.
Validated through high-fidelity simulations, the proposed system demonstrated high mission success rates and accurate formation tracking, confirming the feasibility of safe and efficient natural language-based swarm UAV control.