ABSTRACT To promote the efficient, comprehensive, reliable, and low‐cost testing and application of intelligent algorithms for autonomous underwater vehicles (AUVs), this paper proposes an innovative six‐dimensional digital twin (6D DT) conceptual model and provides detailed engineering implementation strategies of this twin system. This model integrates six core dimensions, including physical entity, virtual entity, virtual native entity (VNE), twin data, services, and communication connection. The concept of VNE is introduced to significantly enhance the practicability, security, and reliability of AUV testing by constructing diversified test scenarios. To implement the proposed model, a high‐fidelity underwater Cyberspace visualization is developed using Unreal Engine 5, which improves the granularity of virtual–real mapping and enhances human–computer interaction. An efficient data bridge plugin is implemented to ensure real‐time, stable bidirectional communication. The DT system (DTS) supports both offline simulation and online DT modes, enabling flexible testing from pure software simulation to real‐time virtual–physical interaction, thereby enhancing the credibility of algorithm validation. Two experimental cases conducted on this DTS demonstrate the technical feasibility and reliability of the proposed conceptual model. The approach provides a valuable reference for applying digital twin technology in underwater unmanned systems and accelerates the development of autonomous intelligent AUVs.
Yu et al. (Tue,) studied this question.