LLM-Guided Distributed Model Predictive Control for Decentralized UAV Formations
Bibliographic record
Abstract
Real-time autonomous control of decentralized drone swarms in dynamic and cluttered environments remains a significant challenge. This paper presents a natural language-driven framework that integrates a fine-tuned large language model (LLM) with distributed model predictive control (MPC) to enable scalable and responsive UAV swarm autonomy. The system architecture comprises a ground control unit, an intelligent mission planning agent, and a decentralized swarm of drones. Mission objectives and target coordinates supplied by external sources (e.g., satellites, command center or airborne platforms), are processed by fine-tuned Phi-2 LLM trained on over 200,000 command variations. The LLM interprets these natural language inputs into structured mission plans, including drone assignments, formations, and operational modes (e.g., swarm-based, multi-target, or single-agent deployments). These plans are dispatched via the Agent mission allocator to the UAVs, each of which leverages a local MPC controller to execute its assigned task. The controllers dynamically optimize flight trajectories while ensuring collision avoidance, formation maintenance, and seamless role transitions. The framework is validated in a high-fidelity simulation environment that combines the ROTORS quadrotor dynamics simulator with Unreal Engine’s photorealistic and depth-aware rendering, facilitating vision-based navigation in cluttered environments. Experimental results demonstrate high mission success rates, accurate formation tracking, and robust adaptability to mid-mission updates, affirming the potential of combining LLM-driven intent parsing with decentralized MPC for intuitive, safe, and scalable swarm control. Future work will focus on extending this framework to physical UAV platforms for real-world deployment.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".