06.2 - Flight Dynamics and Control (UAV related)
INFORMATIVE PATH PLANNING BASED ON REDUCED-RANK GAUSSIAN PROCESS FOR HAZARDOUS SUBSTANCE MONITORING
M. Lee¹, M. Park¹; ¹Changwon National Unviersity, South Korea
This study proposes an integrated framework that simultaneously performs information-theoretic, real-time path planning for an unmanned aerial vehicle (UAV) and hazardous substance mapping in environments where hazardous substance are released from within multiple chimneys (cylindrical structures). Considering the characteristics of mobile sensor data suitable for UAV deployment, we propose an Extended Reduced-Rank Gaussian Process (ERRGP) designed for non-Gaussian measurement modeling and applicable to various types of sensors. For path planning, the framework utilizes LiDAR-based frontiers to generate trajectories that actively avoid obstacles while maximizing information gain. The proposed framework is validated through simulation experiments.
DOI: 10.71945/icas2026_0513