Reproduce “Adaptive Tracking and Perching for Quadrotor in Dynamic Scenarios” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yuman Gao, Jialin Ji, Qianhao Wang, Rui Jin, Yi Lin, Zhimeng Shang, Yanjun Cao, Shaojie Shen and 2 more
Researchers from Zhejiang University and DJI developed an adaptive quadrotor system that performs real-time, visibility-aware tracking and perching on moving, inclined platforms. The system, deployed on a commercial DJI MAVIC3 drone, achieved successful perching at speeds up to 30 km/h and on surfaces inclined up to 60 degrees in real-world conditions.
Key insights
- Comprehensive 3D visibility metrics (observation distance, angle, and elastic occlusion effect avoidance) are crucial for robust tracking and occlusion management in dynamic and cluttered environments.
- Flexible terminal constraint strategies, particularly the tangential velocity relaxation, are essential for generating feasible and safe perching trajectories when dealing with the tight, coupled constraints of dynamic, inclined target surfaces.
- The use of efficient spatio-temporal trajectory representation (MINCO) and gradient-based optimization enables real-time, onboard replanning on commercial, resource-limited drones, transitioning complex planning from offboard to deployable hardware.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.