Search-based Motion Planning for Aggressive Flight in SE(3) (1710.02748v1)

Abstract: Quadrotors with large thrust-to-weight ratios are able to track aggressive trajectories with sharp turns and high accelerations. In this work, we develop a search-based trajectory planning approach that exploits the quadrotor maneuverability to generate sequences of motion primitives in cluttered environments. We model the quadrotor body as an ellipsoid and compute its flight attitude along trajectories in order to check for collisions against obstacles. The ellipsoid model allows the quadrotor to pass through gaps that are smaller than its diameter with non-zero pitch or roll angles. Without any prior information about the location of gaps and associated attitude constraints, our algorithm is able to find a safe and optimal trajectory that guides the robot to its goal as fast as possible. To accelerate planning, we first perform a lower dimensional search and use it as a heuristic to guide the generation of a final dynamically feasible trajectory. We analyze critical discretization parameters of motion primitive planning and demonstrate the feasibility of the generated trajectories in various simulations and real-world experiments.

• The paper presents a search-based planning framework that optimizes quadrotor trajectories using motion primitives and ellipsoid modeling.
• It incorporates dynamical constraints like maximum velocity, acceleration, and jerk to attain collision-free, aggressive maneuvers in SE(3) space.
• A hierarchical refinement process is employed to significantly reduce computation time, with both simulation and real-world experiments validating the method.

Search-Based Motion Planning for Aggressive Flight in SE(3)

The paper presents a search-based trajectory planning framework specifically designed to harness the maneuverability of quadrotors in complex, cluttered environments. By modeling the quadrotor as an ellipsoid and considering its flight attitude, the authors aim to generate dynamic yet collision-free trajectories that maximize the potential of quadrotors with high thrust-to-weight ratios. The proposed methodology enables these aerial vehicles to traverse narrow gaps between obstacles, offering innovative insights into motion planning with considerations for attitude constraints.

To tackle the complexities of planning in SE(3) space, where the quadrotor's rotation and translation are intertwined due to its under-actuated nature, the paper builds upon a search-based approach leveraging motion primitives. These primitives are generated via optimal control techniques and are explored to construct a trajectory from an initial state to the goal state. Notably, the quadrotor's body is modeled as an ellipsoid, permitting non-zero pitch or roll angles to navigate through openings potentially narrower than its physical diameter, thus facilitating aggressive flight maneuvers.

Incorporating dynamical constraints such as maximum velocity, acceleration, and jerk, the algorithm checks the feasibility of each motion primitive. The use of ellipsoid modeling offers a notable advancement, circumventing the conservative limitations imposed by spherical approximations typical in collision checking. By systematically addressing and integrating these factors, the paper contributes a significant methodological approach to planning agile quadrotor trajectories in cluttered spaces.

Moreover, the authors propose a hierarchical trajectory refinement process, optimizing planning efficiency by utilizing preliminary trajectories in lower dimensional spaces as heuristics to guide higher dimensional pathfinding. The hierarchical refinement process demonstrates substantial reductions in computation time by confining exploration to regions proximal to the prior trajectory, enhancing the algorithm's applicability in time-sensitive settings.

The paper's simulation and real-world experiments underscore the robustness of the proposed method, successfully executing trajectories in obstacle-dense environments where conventional techniques would falter due to inflated obstacle maps. These experiments substantiate the planner's capability to enable quadrotors to pass through gaps with intricate non-zero roll and pitch angles, showcasing practical applicability in real-world scenarios.

The authors also offer an open-source implementation of their algorithm, promoting accessibility and further research in motion planning methodologies. The detailed examination of parameter effects on trajectory generation provides valuable guidance for optimizing planner settings, ensuring smooth and efficient operations tailored to specific mission profiles.

In summary, this work represents a pivotal advancement in search-based motion planning for quadrotors, delivering crucial insights into handling SE(3) planning challenges and showcasing a practical, effective framework for leveraging the full potential of aerial vehicle agility in navigation tasks amidst substantial environmental constraints. The methodological contributions hold promising implications for future developments in AI-driven flight dynamics and autonomous navigation systems.