HEPP: Hyper-efficient Perception and Planning for High-speed Obstacle Avoidance of UAVs

TL;DR

HEPP addresses high-speed obstacle avoidance for UAVs in cluttered environments by integrating an incremental robocentric map with direct distance and gradient information (RESDF), an obstacle-aware topological path search that yields multiple diverse routes, and an adaptive gradient-based trajectory optimization with adaptive initialization. The system achieves real-time latency of a few milliseconds and supports speeds near 15 m/s in challenging environments, as validated in both simulations and real-world tests. The key contributions are the Incremental Robocentric Point Cloud Map with RESDF, the obstacle-aware topology search, and the MINCO-based adaptive trajectory initialization and optimization framework, enabling near-global-optimal performance. This work advances practical high-speed UAV navigation by delivering a robust perception-planning pipeline suitable for dense, unknown environments.

Abstract

High-speed obstacle avoidance of uncrewed aerial vehicles (UAVs) in cluttered environments is a significant challenge. Existing UAV planning and obstacle avoidance systems can only fly at moderate speeds or at high speeds over empty or sparse fields. In this article, we propose a hyper-efficient perception and planning system for the high-speed obstacle avoidance of UAVs. The system mainly consists of three modules: 1) A novel incremental robocentric mapping method with distance and gradient information, which takes 89.5% less time compared to existing methods. 2) A novel obstacle-aware topological path search method that generates multiple distinct paths. 3) An adaptive gradient-based high-speed trajectory generation method with a novel time pre-allocation algorithm. With these innovations, the system has an excellent real-time performance with only milliseconds latency in each iteration, taking 79.24% less time than existing methods at high speeds (15 m/s in cluttered environments), allowing UAVs to fly swiftly and avoid obstacles in cluttered environments. The planned trajectory of the UAV is close to the global optimum in both temporal and spatial domains. Finally, extensive validations in both simulation and real-world experiments demonstrate the effectiveness of our proposed system for high-speed navigation in cluttered environments.

Figures (19)

• Figure 1: A demonstration of high-speed navigation, the UAV flies at over 11.0 m/s across the forest.
• Figure 2: An overview of our system framework.
• Figure 3: The illustration of the map sliding process. The points in the old map range that do not intersect with the current map range are deleted to release the memory usage.
• Figure 4: The illustration of the raycasting process. The point will be removed if it is passed by a new laser scan, where the noise and the dynamic points can be eliminated.
• Figure 5: The comparison of traditional ESDF (a) and our RESDF (b). ESDF needs to be updated in the whole map, and our method can obtain the distance and gradient directly.