Overview of Research on Transformation of Multi-AUV Formations

Abstract

Because of their wide detection range and rich functions, autonomous underwater vehicles (AUVs) are widely used for observing the marine environment, for exploring natural resources, for security and defense purposes, and in many other fields of interest. Compared with a single AUV, a multi-AUV formation can better perform various tasks and adapt to complex underwater environments. With changes in the mission or environment, a change in the UAV formation may also be required. In the last decade, much progress has been made in the transformation of multi-AUV formations. In this paper, we aim to analyze the core concepts of multi-AUV formation transformation; summarize the effects of the AUV model, underwater environment, and communication between AUVs within formations on formation transformation; and elaborate on basic theories and implementation approaches for multi-AUV formation transformation. Moreover, this overview includes a bibliometric analysis of the related literature from multiple perspectives. Finally, some challenging issues and future research directions for multi-AUV formation transformation are highlighted.

Artificial intelligence scheduling and task planning for collaborative operations of remote-controlled submarines and manned submarines

“…In the task planning section, a task planning method based on event driven and state dependent is proposed, taking into account the characteristics of underwater robot operations. Firstly, this article models the operational behavior between RCS and manned submersibles as events [14][15]. Then, based on the analysis of task constraints and robot capabilities, the task was decomposed and an event driven task planning method based on Dijkstra algorithm and A * algorithm was designed.…”

From animal collective behaviors to swarm robotic cooperation

“…Meanwhile, it is difficult for one single UMV to conduct complicated tasks. However, the AUV swarm could implement tasks cooperatively to achieve greater efficiency in the complex marine environment [ 89 ]. Thus, multi-UUV cooperation will improve the intelligence and efficiency of multiple individuals performing tasks independently, enabling them to better accomplish tasks that cannot be conducted by one individual.…”

Multivariate linear regression model based on cross-entropy for estimating disorganisation in drone formations

“…The avoidance of collisions with intruders was a widely discussed issue in relation to drones and their formations. For example, models of vehicle formation control 20 or formation control must allowed drones to reorganised and changed their positions in order to perform specific tasks 21 , or must position robots at specific positions over time 22 . However, the additional effort (and hence energy) that the swarm must consumed in order to avoided the intruder tend to be ignored, as research in this area focused on the efficiency of the collision-avoidance algorithm.…”