Global-State-Free Obstacle Avoidance for Quadrotor Control in Air-Ground Cooperation
Baozhe Zhang 2 2 , Xinwei Chen 2 1, 2 , Qingcheng Chen 3 , Chao Xu 1, 2 , Fei Gao 1, 2 , and Yanjun Cao 1, 2 · 2025
Read Global-State-Free Obstacle Avoidance for Quadrotor Control in Air-Ground Cooperation by Baozhe Zhang 2 2 , Xinwei Chen 2 1, 2 , Qingcheng Chen 3 , Chao Xu 1, 2 , Fei Gao 1, 2 , and Yanjun Cao 1, 2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-24315.txt.
Introduction
Baozhe Zhang and five co-authors add obstacle avoidance to a control framework for cooperating drones and ground vehicles that works without global position estimates. The base framework, CoNi-MPC, uses only relative states between the air and ground robots, which is efficient but carries no information about obstacles. Their new algorithm, CoNi-OA, fills that gap.
CoNi-OA uses a single frame of raw LiDAR data from the drone to build a modulation matrix. That matrix directly adjusts the quadrotor's velocity to steer around obstacles in real time. The method needs no global state and no prediction of obstacle motion, suiting fast air ground cooperation in cluttered settings.
Catalog record
| CREATOR | Baozhe Zhang 2 2 , Xinwei Chen 2 1, 2 , Qingcheng Chen 3 , Chao Xu 1, 2 , Fei Gao 1, 2 , and Yanjun Cao 1, 2 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-24315
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Global-State-Free Obstacle Avoidance for Quadrotor Control in Air-Ground Cooperation Baozhe Zhang 2 2 , Xinwei Chen 2 1, 2 , Qingcheng Chen 3 , Chao Xu 1, 2 , Fei Gao 1, 2 , and Yanjun Cao 1, 2 Abstract CoNi-MPC [ 1 ] 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, Co operative N on- i nertial frame-based O bstacle A voidance (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. \textsuperscript{2} \textsuperscript{2} footnotetext: Equal contribution † † footnotetext: This work was supported by National Nature Science Foundation of China under Grant 62103368. (Corresponding author: Yanjun Cao, Chao Xu. {yanjunhi, cxu}@zju.edu.cn) \textsuperscript{1} \textsuperscript{1} footnotetext: State Key Laboratory of Industrial Control Technology, Institute of Cyber-Systems and Control, Zhejiang University, Hangzhou, 310027, China. \textsuperscript{2} \textsuperscript{2} footnotetext: Huzhou Institute of Zhejiang University, Huzhou, 313000, China. \textsuperscript{3} \textsuperscript{3} footnotetext: Shanghai Institute of Special Equipment Inspection & Technical Research, Shanghai, 200062, China. I Introduction Multi-robot cooperation systems, such as UAV-UGV pairs [ 2 ] , leader-follower systems [ 3 ] , multi-agent formations [ 4 ] , and autonomous landing platforms [ 5 ] have been extensively utilized in surveillance, search-and-rescue operations, and cinematography. In CoNi-MPC [ 1 ] , the authors proposed an air-ground robot system in which the UAV (or quadrotor) is actively controlled to fulfill a task along with an independently controlled UGV. It eliminates the need for SLAM and continuous trajectory re-planning, offering a computationally efficient and robust approach that adapts seamlessly to dynamic and unpredictable environments, making it broadly suitable for diverse air-ground cooperative applications. However, CoNi-MPC does not consider the influence of obstacles in the environment, and the UAV cannot guarantee its own safety during flight, strongly limiting its application. In this letter, we propose a novel obstacle avoidance method, Co operative N on- i nertial frame based O bstacle A voidance (CoNi-OA) for quadrotor control, to address the safety issue in UAV-UGV collaboration system. Fig. 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. Directly controlling the UAV in the UGV’s non-inertial frame provides a more intuitive and efficient way to perform air-ground cooperative tasks by removing the dependence on the global state of the UAV in the world frame. Within this cooperative framework, directly controlling the UAV enables it to execute cooperative tasks such as leader-follower control, directional landing, and orbit flight, without the need to both predict UGV motion and continuously replan global trajectories. Considering that the global state estimation in the world frame depends on either extra infrastructure (GPS, motion capture systems, etc.) or featured environments (visual or geometry features, etc.), control in the UGV’s non-inertial frame greatly expands the application scenario to feature-less or unknown environments. However, it also introduces new challenges for obstacle avoidance function which is unavoidable in real-world applications. For example, the UAV must avoid obstacles in the UGV’s non-inertial reference frame, and the static obstacles in the inertial world reference frame become “ dynamic ’’ 1 1 1 In this work, we define “dynamic” obstacles as obstacles whose positions change unpredictably in the UGV’s non-inertial frame due to the UGV’s unknown motion. Conversely, obstacles that remain fixed relative to the UGV are considered “static”. and unpredictable in the non-inertial frame due to the unknown UGV’s movements. Therefore the UAV must be able to avoid both static and dynamic obstacles in real time while maintaining its dynamic feasibility in the non-inertial frame. There are three main challenges to address in this context: • Avoiding unpredictable dynamic obstacles : Obstacles in the non-inertial frame become dynamic and unpredictable due to the unknown UGV’s motion. The UAV need to effectively avoid these dynamic obstacles while following the planned relative motion to the UGV. • Rapid response capability : Achieving rapid obstacle avoidance is crucial, particularly for environments with dynamic obstacles. Methods based on real-time optimization often rely on static assumptions like Euclidean Signed Distance Fields (ESDFs) or flying corridors, where the update of them can lag in dynamic environments. Delays in updating these ESDFs or constraints hinder their responsiveness and may lead to collisions. • Quadrotor dynamic feasibility : Ensuring that the generated trajectories comply with the UAV’s dynamic constraints is essential for safe and smooth flight. This is particularly challenging in the non-inertial frame for air-ground cooperation, where the UAV must account for the influence of unknown UGV motion on its own dynamics. To overcome these challenges, we propose CoNi-OA method as a local planner for the quadrotor with a LiDAR sensor mounted on it to avoid dynamic obstacles in the UGV’s non-inertial frame with only one frame of LiDAR sampling in each control iteration. Inspired by [ 6 ] , we employ a sample-based modulation matrix (directly from LiDAR sampled data points) to modulate the initial velocity from a high-level planner, iteratively generating local collision-free trajectories for the quadrotor that adhere to the non-inertial dynamical system (DS) of the quadrotor. Different from [ 6 ] that applying the modulated DS to a point-mass model in 2D space, CoNi-OA is designed for a rigid body like a quadrotor in 3D space and in the non-inertial frame, which supports more aggressive maneuvers than the point-mass model. The generated trajectories are then effectively tracked by CoNi-MPC, enabling the quadrotor to cooperatively avoid obstacles in dynamic scenarios. Novelty and Contribution The contributions of this paper are summarized as follows: • First non-inertial frame obstacle avoidance for UAV-UGV cooperation : Unlike existing methods requiring global states or planning in the world frame [ 7 , 8 , 9 ] , we propose the first modulation-based solution specifically designed for UAV control in a UGV’s non-inertial frame. This eliminates dependency on global positioning systems or SLAM while handling both static and dynamic obstacles through a LiDAR sensor. • Single-frame reactive modulation : Compared to optimization-based approaches [ 7 ] that require obstacle modeling/prediction, our method directly processes raw LiDAR points to compute velocity modulation matrices in <5 ms per