Study on Classification and Detection of Operational Fatigue in Deep‐Well Mining Truck Drivers Based on GSWOA‐LSSVM
Year: 2026
Type: article
Abstract: ABSTRACT Effectively detecting the operational fatigue of miners can identify safety risks in mines. As the executors of auxiliary transportation in mines, the operational fatigue status of mining truck drivers deserves particular attention and detection. To reduce safety production accidents caused by the operational fatigue of mining truck drivers, this study conducted simulated loading, hauling, and unloading operation experiments for deep‐well mining truck drivers. Functional near‐infrared (f_NIRS) brain data were dynamically collected from 24 mining truck drivers before and after their night‐shift operations. Combining principal component analysis (PCA), three characteristic indicators were selected: the mean, variance, and linear regression slope of oxyhemoglobin (HbO 2 ) in the prefrontal cortex. A classification model for operational fatigue based on the Global Search Strategy‐based Whale Optimization Algorithm‐optimized Least Squares Support Vector Machine (GSWOA‐LSSVM) was established. The research results indicate that there is a significant negative correlation between the HbO 2 concentration in the prefrontal cortex of mining truck drivers and their operational fatigue level ( p < 0.01). Changes in HbO 2 concentration can serve as an objective physiological indicator for detecting the operational fatigue of mining truck drivers. The GSWOA‐LSSVM model achieved an average accuracy of 96.875% in identifying four categories of operational fatigue states: alert, mild, moderate, and severe. This represents an improvement of 14.585% compared to the traditional Support Vector Machine (SVM) model (82.29%) and an improvement of 9.375% compared to the LSSVM model (87.50%). Moreover, its precision, recall, and F 1‐Score are all superior to those of traditional models. Therefore, the GSWOA‐LSSVM classification model is suitable for detecting the operational fatigue of deep‐well mining truck drivers. Accurate detection of the operational fatigue status of deep‐well mining truck drivers can provide theoretical support and practical guidance for the construction of safety management systems, such as major accident prevention and control and on‐site operation supervision.
Source: Safety Science and Technology
DOI: 10.1002/sst3.70036
Authors: Ying Chen, Jikun Lu, Caiwu Lu, Qinghua Gu, Song Jiang
Institution: Xi'an University of Architecture and Technology
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7168026572