智能印刷中的实时注意力监测：一种面向工业5.0的人本化EEG谱熵方法

Real-Time Attention Monitoring in Smart Printing: A Human-Centred EEG Spectral Entropy Approach for Industry 5.0

摘要: 在工业5.0人本范式下，精密印刷质量控制亟需对操作人员注意力进行有效监测。本文提出一种基于实时EEG的注意力评估算法，将谱熵与传统频带特征相结合表征认知状态。基于同一脑电数据条件下的对比实验结果表明，所提出方法在注意力评估中相较传统频段功率指标表现出更高的稳定性与一致性，其输出差异在统计意义上达到显著水平，同时保持较低的计算成本与良好的可解释性。研究结果验证了谱熵特征在工业脑电注意力监测中的工程有效性，为工业5.0背景下将人因因素融入智能制造系统提供了可行的人本化技术路径。

Abstract: Under the human-centred paradigm of Industry 5.0, precision printing quality control urgently requires effective monitoring of operator attention. This paper proposes an attention assessment algorithm based on real-time EEG, combining spectral entropy with traditional frequency band features to characterise cognitive states. Comparative experimental results under identical EEG data conditions demonstrate that the proposed method exhibits higher stability and consistency in attention assessment compared to traditional frequency band power metrics. The output differences are statistically significant, while maintaining low computational cost and good interpretability. The findings validate the engineering efficacy of spectral entropy features in industrial EEG attention monitoring, offering a viable human-centred technical pathway for integrating human factors into intelligent manufacturing systems within the Industrial 5.0 framework.