基于多元线性回归的煤矿驾驶员情绪预测模型*

为研究煤矿人车驾驶员情绪状态不良引发的煤矿驾驶事故问题，设计眼动实验并建立基于多元线性回归的驾驶员情绪状况预测模型对驾驶员情绪状况进行预测；通过眼动仪采集煤矿驾驶员良好情绪与不良情绪状态下的各项眼动指标，记录其主观情绪状态；使用多元线性回归方程对数据进行分析与建模，采用平均相对误差对预测模型进行评估。结果表明:每秒注视点个数、平均扫视速度、反应时间、危险源辨识个数在情绪变化前后存在显著差异，且与情绪状态呈较强相关；基于多元线性回归的驾驶员情绪状况预测模型预测精度较高，平均相对误差为8.16%。模型适用于煤矿人车驾驶员的情绪监测，可为煤矿驾驶员安全行驶提供保障。

Prediction model for emotion of coal mine drivers based on multiple linear regression

In order to study the problem of coal mine driving accidents caused by the bad emotional state of coal mine drivers,the eye movement experiment was designed,and a prediction model for the emotional state of drivers based on the multiple linear regression was established to predict the drivers’ emotional state.Various eye movement indexes of coal mine drivers under the good and bad emotional states were collected by the eye tracker,and their subjective emotional states were recorded.The multiple linear regression equations were used to analyze and model the data,and the average relative error was used to evaluate the prediction model.The results showed that the number of fixation points per second,average scanning speed,reaction time and risk identification number were significantly different before and after the emotional change,and were strongly correlated with the emotional state.The prediction model of drivers’ emotional state based on the multiple linear regression had a higher prediction accuracy,with an average relative error of 8.16%.It can be seen that the model is suitable for the emotion monitoring of drivers in the coal mine,and provides guarantee for the safe driving of coal mine drivers.