Fully Automated UAV Cluster Pollution Source Detection Algorithm

In this paper, considering the challenge of gas pollution prevention and the development of unmanned system technology, a type of pollution source detection algorithm which utilizes unmanned aerial vehicle is investigated. The whole research is based on turbulence gas diffusion model derived from Fick’ s laws of diffusion, aiming to determine the location of gas pollution source in a certain area by using drone cluster to measure gas concentration in each sampling point. Firstly, based on the gas diffusion model, this paper proposes a gas pollution source estimation algorithm to provide the target detection for the algorithm. Then, in pursuit of higher execution efficiency and to reduce the risk of collision, an optimized path planning algorithm and an obstacle avoidance algorithm are designed. These algorithms are integrated as a whole, guiding drone cluster to continuously approach the gas pollution sources through multiple rounds of tasks with optimal paths. Finally, simulation is provided to test the whole algorithm, displaying the validity of this algorithm.