Multi-UAV Method for Continuous Source Rate Estimation of Fugitive Gas Emissions from a Point Source

Abstract

A multi unmanned aerial vehicle (UAV) based method for making continuous source rate estimations is proposed. The multi-UAV method utilizes the signal processing, spatial interpolation, and data-driven techniques to apply centroidal Voronoi tessellations (CVT) based coverage control. The method uses low frequency modes from dynamic mode decomposition (DMD) as the CVT density kernel. The method also employs an ordinary kriging estimation variance based entropy as a regularization term to improve exploration behavior. A 3D short time-scale plume simulation model (POSIM) is used to test the continuous source rate estimation performance.