NARA Discovery
Article Details
← Back to Search Results
Journal Article

A Digital Twin Prototype for a Deep-Sea Observation Network: Virtual Environment Reconstruction and Data-Driven Predictive Analytics

Xinya Zhang; Ruixin Chen; Rufu Qin
Journal of Marine Science and Engineering · Vol. 14, Issue 9 · pp. 800 · 2026

Abstract

Effective operation and maintenance (O&M) of deep-sea observation networks are challenged by complex environments and energy limitations. While digital twin (DT) technology offers promising solutions, existing frameworks struggle with high-fidelity, multi-platform orchestration and predictions of electrical energy state. This study proposes a DT framework for a deep-sea observation network (DSON-DT), encompassing telemetry acquisition, predictive analytics, and feedback control to realize a closed-loop workflow for monitoring and managing platform states within virtual scenes. Powered by real-time Internet of underwater things (IoUT) data, a high-fidelity virtual environment is constructed in the Unreal Engine 5 game engine, accurately mapping ambient marine environments and reconstructing platform dynamic behaviors via data-driven approaches and geometric constraints. An improved auto-regressive long short-term memory (AR-LSTM) network is proposed to forecast the battery state of charge (SoC). Experimental results show that this algorithm effectively mitigates the impacts of severe deep-sea noise and the flat open-circuit voltage plateau, suppressing state oscillations to provide reliable references for proactive endurance management. The Vue.js-based web prototype, deployed via pixel streaming, offers seamless interfaces for interactive visualization, analysis, and remote operation. This research achieves comprehensive situational awareness for deep-sea platforms, providing validated technical support for the holistic evaluation and intelligent O&M of heterogeneous marine infrastructures.

Bibliographic Information

JournalJournal of Marine Science and Engineering
PublisherMDPI
Publication Date2026-04-27
Publication Year2026
Volume14
Issue9
Pages800
Document TypeJournal Article
eISSN2077-1312
DOI10.3390/jmse14090800
SubjectMarine science; oceanography; marine engineering; coastal science; marine environment

Access Information

NARA Access CoverageOA / free full text
Journal Homepagehttps://www.mdpi.com/journal/jmse
Publisher PageOpen Publisher Page
This article is openly available from the publisher.