Aerial Plume Localization and Tracking for Fixed-Wing Glider Sensor Platforms

Unpowered, expendable gliding aircraft that can be aerially deployed in large numbers offer flexible atmospheric sampling capabilities in dangerous areas at the cost of inherently limited mission durations. This paper describes a glider guidance strategy that uses in-flight measurements to estimate and steer the aircraft towards the center of pollutant plumes in order to improve sensor placement during flight. This concept was implemented within a mission simulation environment in which gliders used onboard sensor measurements in conjunction with a reduced-order, quasi-steady Gaussian plume model to localize the source of an airborne pollutant using Bayesian techniques. A prototype glider airframe and an autopilot package were developed and flown through a series of closed-loop aerial-drop tests to establish the performance parameters of a representative platform for use in these simulations. Numerical results indicate that the proposed estimation strategy continued to provide useful position estimates of the plume source when a significant, synthetic measurement noise is injected. A first-order guidance strategy developed in conjunction with the estimation scheme has also been shown to successfully direct aircraft towards likely pollutant sources through simple flight directives that can be executed by a rudimentary gliding aircraft for atmospheric sampling.