Hybrid brain-machine interface driven rehabilitation assistive robot
By combining a dynamic spatiotemporal graph model with multimodal perception and common sense knowledge fusion, the shortcomings of existing rehabilitation robots in terms of interaction reliability, environmental adaptability and operational precision are solved. It realizes intelligent user intent interpretation and environmental cognition, and improves the safety and intelligence of rehabilitation assistive robots.
Patent Information
- Application Number
- CN202610752258.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing rehabilitation robots have deficiencies in terms of interaction reliability, environmental adaptability, decision-making intelligence, and operational precision. They cannot effectively distinguish between conscious control and unconscious physiological activities, lack environmental contextual understanding, lack a unified state model, lack guidance from physical common sense during operation, and provide crude and unsafe feedback.
Using a dynamic spatiotemporal graph as a unified cognitive model, the system collects signals through a multimodal perception module, constructs and updates the dynamic spatiotemporal graph, and combines a common sense knowledge fusion module and graph neural network reasoning to achieve intelligent parsing of user intent, robust cognition of the environment, and adaptive and precise execution of operations.
It improves the reliability and naturalness of interaction, achieves stable target perception and positioning in complex environments, ensures motion safety, enables safe and stable grasping of diverse objects, and reduces user cognitive load and operational fatigue.
Smart Images
Figure CN122398595A_ABST