An Underwater Glider Fleet Coordinate Control Architecture for Ocean of Things to Enable Transparent Ocean

Ocean of Things (OoT) is a novel technology that integrates marine devices and the Internet of Things, which is the basis for realizing a transparent ocean. However, due to the uncertainty of the marine environment and the constraint of underwater communication, there are still some challenges to building OoT in a large area. Therefore, this article proposes an underwater glider (UG) fleet coordinate control architecture (GFC2A) to realize large-range ocean observation based on real-time data and ocean model data. First, a data-driven velocity and time prediction model (VTPM) for UG is put forward with a novel method, principal component analysis combined with genetic algorithm and back propagation neuralnetwork (PCA-GA-BPNN), and its accuracy is compared with that of other benchmark methods, including the dynamic model. Second, to realize the coordinate control of the UG fleet, a novel virtual distributed control model is built on the dynamic leader–follower method and VTPM, which can predict the position of a UG underwater with VTPM and synchronize that with the surface UG. Then, the GFC2A is developed by integrating a virtual distributed control model, ocean model data, and database, and then verified by a simulation experiment and two sea trials carried out in the northern South China Sea. According to the results, the mean squared errors of velocity and time in VTPM are 0.0347 m/s and 2.5 × 10
⁻⁴
min, respectively. The mean deviation of fleet formation is about 1.39 km, and the time synchronization error is limited to 10 min. The GFC2A proposed in this article also provides a reference for the cooperation work of unmanned or manned vehicles in a communication-constrained environment.