Design of a Distributed Controller for Model-Augmented Gas Distribution Mapping with Robotic Swarms

Kurzfassung

Atmospheric gas plumes are a highly dynamic spatial process, which makes it difficult to map their distributions. Robotic swarms equipped with gas sensors are well-suited for this task, especially when their control strategy exploits knowledge of the underlying physics. In this paper, a distributed control framework for a swarm system is developed that can track the contour of a gas plume in a dynamic wind field. As a result, agents contract around an isoline of gas concentration, which is then utilized to estimate the spatial gas distribution. Simulation studies show that the distributed control scheme is capable of following the dynamic evolution of the gas dispersion process and can accurately reconstruct the gas concentration field and even its source. In comparison to benchmark strategies, this method improves estimation accuracy. The core innovation of this work lies in addressing gas distribution mapping through the intermediate objective of contour line tracking. The proposed distributed controller effec- tively locates and tracks gas plumes while estimating the corresponding concentration fields.