In order to ensure the order picking efficiency and the efficient operation of warehousing, and to reduce the mental fatigue caused by physical fatigue under human-computer interaction the changing rules and characteristics of mental fatigue of operators studied. Twenty-one male participants were required to complete three order selection tasks with different levels of physical fatigue, and their subjective questionnaires and electroencephalogram(EEG) signals were collected during the task. The results show θ，（α+β）/θ of frontal and occipital regions can be used as typical parameters to characterize degree of mental fatigue. Meanwhile, the machine learning classification method based on autoregressive model and support vector machine can quickly evaluate mental fatigue. This method owns fast computation speed and accuracy. The results of this study can provide a basis for the formulation of comprehensive and differential fatigue management measures for intelligent warehousing logistics picking operation under human-computer interaction.

为了保证智能仓储系统订单拣选效率及仓储的高效运行，减少人机交互下操作者的体力疲劳诱发的脑力疲劳，需要对操作者的脑力疲劳变化规律和变化特征进行研究。通过21名男性参与者被要求完成三个不同体力疲劳下的订单拣选任务，并在任务期间收集他们的主观问卷和脑电信号进行研究。结果表明高体力疲劳能显著加速脑力疲劳的诱发，额区和枕区α、β节律和β/θ和(α+β)/θ等参数可以作为典型的表征脑力疲劳的参数。同时基于自回归模型和支持向量机的机器学习分类方法可以快速地对脑力疲劳进行评价，该方法具有运算速度快和准确率的特征。研究结果可为人-机交互下智能仓储物流拣选操作现场的全面性、差异化疲劳管理措施的制定提供依据。

北京市教育委员会科学研究计划(KM202210037001)