Research on classification and determination method of work fatigue level of coal mine dispatchers

Abstract

In order to accurately determine the level of fatigue of coal mine monitoring and dispatching operators and to reduce the error rate of coal mine monitoring and dispatching operations, a simulation experiment of coal mine monitoring and dispatching was carried out using eye tracking technology to collect the operator’s eye movement data and objective and subjective fatigue determination values. The K-means clustering algorithm was used to classify fatigue levels and train a neural network to build a fatigue prediction model for coal mine monitoring and dispatching operations. The results showed that the best fatigue level was classified into three categories and the fitting degree of the neural network prediction model was 90.58%. The prediction model was tested on a coal mine monitoring operation model in Shanxi Province, and the average error of the model field prediction was 6.26%, which was a good prediction effect.

Keywords

• accident prevention
• coal mine safety
• eye tracking technology
• neural network
• k-means clustering