Global-State-Free Obstacle Avoidance for Quadrotor Control in Air-Ground Cooperation

TL;DR

This work tackles obstacle avoidance for UAVs operating in the UGV's non-inertial frame by introducing CoNi-OA, a modulation-based local planner that relies on a single LiDAR frame and eliminates the need for global state estimation or obstacle prediction. It builds a 3D extension of a fast, sample-based modulation of a dynamical system, computing a modulation matrix ${\boldsymbol{\mathcal{M}}}$ from LiDAR data and using it to generate collision-free reference trajectories that CoNi-MPC can track in real time. Key contributions include the first non-inertial-frame obstacle avoidance for UAV-UGV cooperation, single-frame reactive modulation with $O(nh)$ trajectory generation, and a general framework applicable to static and dynamic environments. The approach demonstrates real-time performance (sub-millisecond loop times for planning and sub-5 ms trajectory generation) and robust obstacle avoidance in both simulations and real-world tests, enabling safe, dynamic air-ground cooperation in featureless or unknown settings.

Abstract

CoNi-MPC provides an efficient framework for UAV control in air-ground cooperative tasks by relying exclusively on relative states, eliminating the need for global state estimation. However, its lack of environmental information poses significant challenges for obstacle avoidance. To address this issue, we propose a novel obstacle avoidance algorithm, Cooperative Non-inertial frame-based Obstacle Avoidance (CoNi-OA), designed explicitly for UAV-UGV cooperative scenarios without reliance on global state estimation or obstacle prediction. CoNi-OA uniquely utilizes a single frame of raw LiDAR data from the UAV to generate a modulation matrix, which directly adjusts the quadrotor's velocity to achieve obstacle avoidance. This modulation-based method enables real-time generation of collision-free trajectories within the UGV's non-inertial frame, significantly reducing computational demands (less than 5 ms per iteration) while maintaining safety in dynamic and unpredictable environments. The key contributions of this work include: (1) a modulation-based obstacle avoidance algorithm specifically tailored for UAV-UGV cooperation in non-inertial frames without global states; (2) rapid, real-time trajectory generation based solely on single-frame LiDAR data, removing the need for obstacle modeling or prediction; and (3) adaptability to both static and dynamic environments, thus extending applicability to featureless or unknown scenarios.

Figures (10)

• Figure 1: A quadrotor avoids obstacles and lands on a rotating UGV from a fixed direction by applying CoNi-OA, in which the pre-generated collision-free trajectory (green) is obstructed during landing. (a) shows the experiment from a third-person view in the world frame where the quadrotor avoides the "dynamic" obstacle in the UGV's non-inertial frame; (b) shows the final successful landing.
• Figure 2: The point-mass model's collision-free navigation around obstacles using the modulated DS. Red arrow indicate the summed reference direction from obstacles at position $\boldsymbol{\mathbf{\xi}}$, with the purple and orange arrows showing the direction of the initial and modulated velocities, respectively.
• Figure 3: CoNi-OA adds obstacle avoidance to CoNi-MPCzhang2023coni for robust UAV-UGV collaboration. (a) The UAV, operating in the UGV's non-inertial reference frame with pre-computed task trajectory which can be blocked by dynamic obstacles, uses lidar data to compute safe trajectories while adapting to the UGV's unpredictable motion. (b) By modulating the initial velocity on the pre-computed task trajectory, the UAV generates collision-free trajectories toward the original task trajectory. (c) shows the labels.
• Figure 4: Visualization of $\lambda_r$ and $\lambda_t$ with respect to $\lVert \boldsymbol{\mathbf{r}}(\xi)\rVert$.
• Figure 5: System overview and implementation of UAV-UGV cooperative obstacle avoidance using CoNi-OA.