Advances in vital‐sign prediction and early‐warning models for underground coal mine workers integrating environmental factors

Abstract Coal mining faces, as high‐risk operational environments, present a severe threat to miners' occupational health due to the complex, dynamic coupling of multiple factors such as dust, temperature, humidity, noise, and harmful gases. This paper systematically reviews research advances in environmental parameter monitoring and personnel vital‐sign detection methods. It synthesizes the mechanisms underlying the nonlinear impacts of environmental stressors—including dust exposure and high‐temperature, high‐humidity conditions—on miners' physiological systems (e.g., cardiovascular and respiratory functions). Furthermore, it consolidates technical pathways for multi‐source data fusion‐based early‐warning models. Current research demonstrates that data preprocessing methods leveraging adaptive threshold filtering and GANs significantly enhance data quality. The synergistic application of statistical thresholds, machine learning, and digital‐twin technologies offers novel approaches for dynamic environment–physiology early warning. However, persistent challenges remain, including insufficient standardization of environment–physiology data, weak cross‐scenario model generalizability, and difficulties in quantifying multi‐factor coupling effects. Future research should establish interdisciplinary frameworks integrating occupational medicine and engineering principles, develop context‐adaptive early‐warning systems, and advance coal mine safety management toward a closed‐loop “monitoring–warning–intervention” paradigm. Through a critical synthesis of existing achievements and limitations, this study provides theoretical insights for mitigating occupational health risks in coal mines and facilitates the evolution of safety monitoring from single‐hazard alerts toward multidimensional health surveillance.