论文部分内容阅读
对于航空训练中飞行员工作负荷状态的识别,可以有效的确保航空安全。通过飞行数据来对飞行员工作负荷识别,需要对飞行数据进行提取,根据飞行数据特点,建立飞行特征集,完成对负荷状态的识别。传统方法通过建立人体工作负荷状态评估模型对工作状态进行识别,但无法得到飞行特征集,导致识别精度低。提出了基于Treelets降维的飞行员工作负荷状态智能识别方法。首先引入时域信号特征均值、方差与均方根对飞行数据进行提取,根据飞行数据特点,建立飞行特征集,利用飞行特征集组建Treelets降维算法模型;然后根据布谷鸟算法优化的高斯分类模型,评估飞行员工作负荷状态;最后将结果与NASA-TLX的飞行员工作负荷结果对比,利用比对结果完成对航空训练中飞行员工作负荷状态的智能识别。实验结果表明,所提方法能够实现对飞行员的工作状态负荷状态的有效识别,且识别精度较高。
Aviation pilot training for the identification of the workload of the state, can effectively ensure aviation safety. Through the flight data to the pilot workload identification, the need to extract the flight data, according to the characteristics of flight data, the establishment of flight feature set to complete the load status identification. The traditional method can identify the working status by establishing the model of human workload assessment, but can not obtain the flight feature set, resulting in low recognition accuracy. A method of intelligent recognition of pilot workload based on Treelets dimension reduction is proposed. First of all, the time-domain signal features mean, variance and root mean square are used to extract the flight data. According to the characteristics of the flight data, the flight feature set is established and the Treelets dimensionality reduction model is established by using the flight feature set. Then, based on the Gaussian classification model , To assess pilot workload status. Finally, the results are compared with those of NASA-TLX pilots, and the results of the comparison are used to complete the intelligent recognition of pilots workload status in aviation training. The experimental results show that the proposed method can effectively identify the pilots’ working state load status with high recognition accuracy.