一种基于数字孪生的水下机器人智能作业方法与系统
By constructing a digital twin environment and reinforcement learning constrained by physical information neural networks, and combining knowledge graphs for task planning and anomaly diagnosis, the problems of simulation and reality discrepancies and insufficient safety adaptive capabilities in intelligent underwater robot operations have been solved, achieving highly reliable and safe operation in complex underwater environments.
Patent Information
- Application Number
- CN202610883627.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-18
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for intelligent underwater robot operations suffer from problems such as discrepancies between simulation and reality, weak cross-domain generalization ability of strategies, insufficient safety adaptation ability under sparse feedback, and poor interpretability of task planning, resulting in insufficient reliability and safety of intelligent operations in complex underwater environments.
We construct an intelligent underwater robot operation system based on digital twins. By inverting and correcting multimodal sensor data online, we adopt physical information neural network constrained reinforcement learning and combine knowledge graphs for task planning and anomaly diagnosis, so as to achieve closed-loop collaborative evolution of perception, decision-making, execution and learning.
It effectively narrows the gap between simulation and reality, enhances the cross-domain generalization ability of strategies and the interpretability of decisions, ensures safe and autonomous operation in complex underwater environments, and achieves continuous co-evolution and reliability.
Smart Images
Figure CN122411532A_ABST
Abstract
Citation Information
Patent Citations
Autonomous Underwater Vehicle (AUV) motion control method and system based on LLM human-computer interaction framework
CN119717649A