Multi‐UAV active target search based on perceptual adaptation in marine environment
Xiang Li, Hebin Wang, Xingfa Shen · Cognitive Computation and Systems · 2022
Abstract This work considers the problem of target search in marine environment. A perceptual adaptive Gaussian sum cubature Kalman filter is proposed for multi‐target search in non‐linear and non‐Gaussian noise environment. Firstly, the robot collects environmental information, conducts environmental perception, and actively makes decisions. Secondly, the Dynamic Time Warping distance and Frechet distance weighted fusion are used to extract the association knowledge by analysing the trajectories of the underwater and surface targets obtained in the search process. The extracted association knowledge is used in the search process to form an evolving search relationship and increase the search efficiency. Finally, it is applied to realise the search of multiple dynamic targets, including underwater and surface targets, in the non‐Gaussian marine environment.