betaFeedback

WorksComputer Science

Real-Time Air Pollution Monitoring and Tracking via Adaptive Density-Driven Multi-UAV Coordination

Abstract

Accurate and timely air pollution monitoring and tracking are critical for protecting public health and supporting effective environmental management. This paper presents an adaptive Density-Driven Control (D2^2C) framework for coordinating multiple unmanned aerial vehicles (UAVs) to track and map dynamically evolving pollutant distributions. The proposed framework extends the D2^2C strategy to guide UAV trajectories based on predictive pollutant propagation and dynamically updated reference densities. A sensor-driven surrogate plume update mechanism is employed, in which reference samples are propagated using locally sensed concentration gradients and wind estimates through gradient attraction, wind advection, inter-sample repulsion, and stochastic diffusion, enabling UAVs to maintain an accurate spatio-temporal representation of the pollution field. The framework utilizes a coordinated structure where individual UAVs share their local measurements with a central coordinator to maintain a unified set of reference samples, facilitating systematic target allocation and trajectory planning. Simulation results under both mild and strong wind conditions demonstrate that the proposed method significantly outperforms state-of-the-art baselines in terms of Sinkhorn distance. Specifically, in scenarios where the plume drifts beyond a fixed monitoring domain under high-velocity, time-varying wind fields, the tested baseline methods fail to maintain tracking, whereas the proposed D2^2C framework successfully adapts to the evolving distribution. These findings highlight D2^2C as an efficient and resilient solution for atmospheric monitoring in dynamic environments.

Paper details

vlnc:W_g8de2zgq
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
arXiv nonexclusive license
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

Not recorded for this paper

View statement
Looking for related papers…