A team performance dynamic evaluation method and system based on multi-modal physiological data fusion
By fusing multimodal physiological data and deep learning models, combined with dynamic graph convolutional networks, the problems of insufficient multimodal data fusion and lack of dynamic evaluation in team performance evaluation are solved. This enables real-time, objective evaluation and high-precision early warning of team member status, improving the adaptability and accuracy of team performance monitoring.
CN122243285APending Publication Date: 2026-06-19CHINA STATE SHIPBUILDING CORP LTD RESEARCH INSTITUTE 719
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA STATE SHIPBUILDING CORP LTD RESEARCH INSTITUTE 719
- Filing Date
- 2026-03-18
- Publication Date
- 2026-06-19
Smart Images
Figure CN122243285A_ABST
Abstract
This invention discloses a method and system for dynamic team performance evaluation based on multimodal physiological data fusion, aiming to address the problems of single evaluation methods, insufficient dynamism, and inadequate multimodal data fusion in existing technologies. The method synchronously acquires EEG, ECG, and EMG signals of team members through a distributed physiological signal acquisition device, and extracts multimodal physiological features using preprocessing techniques such as wavelet transform and adaptive filtering. Multimodal features are fused using a VGG16 network and a cross-attention mechanism to generate individual cognitive load, psychological stress, and physiological fatigue indices. Furthermore, a multi-role network graph is constructed, and team performance is dynamically aggregated using a graph convolutional network to achieve real-time dynamic evaluation and visualization of team performance.
Need to check novelty before this filing date? Find Prior Art