DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells
Yunwoo Lee, Jungwon Park, H. Jin Kim
AI summary
Problem
Multi-agent aerial target tracking struggles to simultaneously prevent inter-agent collisions and occlusions while maintaining target visibility and adapting to dynamic environments in real time.
Approach
The method introduces time-varying Dynamic Buffered Voronoi Cells and Dynamic Inter-Visibility Cells to enforce safety and visibility constraints, combined with Bernstein polynomial motion primitives and a sample-check-select strategy for fast distributed computation.
Key results
- Dynamic Buffered Voronoi Cell (DBVC) and Dynamic Inter-Visibility Cell (DIVC) for real-time collision and occlusion avoidance
- Integration with Bernstein polynomial motion primitives enabling millisecond-level trajectory computation
- Reduced algorithmic conservativeness yielding higher tracking success rates in complex obstacle environments
- Successful hardware validation with multiple MAVs navigating dozens of obstacles while maintaining target visibility
Why it matters
Enables reliable, real-time cooperative aerial surveillance and cinematography by solving the critical trade-off between safety, visibility, and computational efficiency for multi-drone systems.
Abstract
This letter presents a distributed trajectory planning method for multi-agent aerial tracking. The proposed method uses a Dynamic Buffered Voronoi Cell (DBVC) and a Dynamic Inter-Visibility Cell (DIVC) to formulate the distributed trajectory generation. Specifically, the DBVC and the DIVC are time-variant spaces that prevent mutual collisions and occlusions among agents, while enabling them to maintain suitable distances from the moving target. We combine the DBVC and the DIVC with an efficient Bernstein polynomial motion primitive-based tracking trajectory generation method, which has been refined into a less conservative approach than in our previous work. The proposed algorithm can compute each agent’s trajectory within several milliseconds on an Intel i7 desktop. We validate the tracking performance in challeng- ing scenarios, including environments with dozens of obstacles.