LARGE GMP$^{3}$: Learning-Driven, Bellman-Guided Trajectory Planning for UAVs in Real-Time on SE(3)

2025-09-25

AI Summary

Generating real-time, dynamically feasible, smooth, and collision-free 3D trajectories for UAVs in cluttered, high-dynamic environments remains challenging due to coupled translational and rotational dynamics and stringent safety constraints. Method: This paper proposes GMP³, the first framework to formulate path planning on the SE(3) Lie group manifold—unifying position and orientation modeling and jointly optimizing translational and rotational dynamics. It integrates reinforcement learning with a modified Bellman operator for efficient policy updates and introduces a distributed consensus mechanism to coordinate multi-segment trajectory optimization, enhancing convergence and global performance. Deployment is enabled via MAVLink and the DroneManager ground station. Results: Extensive simulation and real-world indoor flight experiments demonstrate reliable obstacle avoidance under high-dynamic conditions, generation of globally smooth trajectories satisfying full dynamical constraints, and real-time execution on physical UAV platforms.

Technology Category

• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Intelligent Robots: Motion and Path Planning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

We propose $ ext{GMP}^{3}$, a multiphase global path planning framework that generates dynamically feasible three-dimensional trajectories for unmanned aerial vehicles (UAVs) operating in cluttered environments. The framework extends traditional path planning from Euclidean position spaces to the Lie group $mathrm{SE}(3)$, allowing joint learning of translational motion and rotational dynamics. A modified Bellman-based operator is introduced to support reinforcement learning (RL) policy updates while leveraging prior trajectory information for improved convergence. $ ext{GMP}^{3}$ is designed as a distributed framework in which agents influence each other and share policy information along the trajectory: each agent refines its assigned segment and shares with its neighbors via a consensus-based scheme, enabling cooperative policy updates and convergence toward a path shaped globally even under kinematic constraints. We also propose DroneManager, a modular ground control software that interfaces the planner with real UAV platforms via the MAVLink protocol, supporting real-time deployment and feedback. Simulation studies and indoor flight experiments validate the effectiveness of the proposed method in constrained 3D environments, demonstrating reliable obstacle avoidance and smooth, feasible trajectories across both position and orientation. The open-source implementation is available at https://github.com/Domattee/DroneManager

Problem

• Generating dynamically feasible 3D trajectories for UAVs in cluttered environments
• Extending path planning from Euclidean spaces to SE(3) for joint motion learning
• Enabling cooperative policy updates among agents under kinematic constraints

Innovation

• SE(3) Lie group planning for UAV trajectories
• Bellman-based operator for reinforcement learning updates
• Distributed multi-agent framework for cooperative planning