Scholarlink

<p>Dust exposure is an occupational hazard and threat to the health of tunnel construction workers. The use of a disposable facepiece is an effective respiratory protective equipment. Environmental factors in tunnel workplaces, human factors, and the breathing resistance of the disposable facepiece can influence the filtration efficiency (FE) and fit factor (FF), affecting the protection factor (PF)&#xa0;of disposable facepieces ultimately. A prediction model of PF that considers multiple influencing factors is necessary to reduce the risk of respiratory occupational diseases for workers. Thus, a hybrid machine learning model (HMLM) is proposed to predict the FE, FF, and PF of disposable facepieces. The HMLM integrates Support Vector Regression (SVR), K-Nearest Neighbors (KNN), and Decision Tree (TREE), with the weights of the three individual models calculated by using the ‌sum of squared errors (SSE)‌. The performance is evaluated using Root Mean Square Error (RMSE). The dataset for training and testing the model is derived from measurement experiments of FE, FF, and PF in a simulated tunnel working environment. The HMLM demonstrated better prediction performance than three individual models for the FE, FF, and PF of disposable facepieces: HMLM's RMSE is 0.0438 (FE), 0.1784 (FF), and 0.2443 (PF); the RMSE of individual models is FE (SVR: 0.0456, KNN: 0.0740, TREE: 0.0941), FF (SVR: 0.2459, KNN: 0.2535, TREE: 0.1864), PF (SVR: 0.4390, KNN: 0.3013, TREE: 0.3265). Compared with the individual model (SVR), HMLM achieves a percentage improvement of 3.9% (FE), 27.5% (FF), and 44.3% (PF), respectively. The findings mitigate the necessity for additional experimental efforts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of a Novel Hybrid Machine Learning Model for Predicting the Protection Factor of Disposable Facepieces in Tunnels

  • 期刊论文

  • 发表日期：

    2026年1月9日

查看全文
  • Zenghui Liu,
  • Chenchen Sun,
  • Liwei Liu,
  • Zengqing Bai,
  • Xiaobing Yang

摘要

Dust exposure is an occupational hazard and threat to the health of tunnel construction workers. The use of a disposable facepiece is an effective respiratory protective equipment. Environmental factors in tunnel workplaces, human factors, and the breathing resistance of the disposable facepiece can influence the filtration efficiency (FE) and fit factor (FF), affecting the protection factor (PF) of disposable facepieces ultimately. A prediction model of PF that considers multiple influencing factors is necessary to reduce the risk of respiratory occupational diseases for workers. Thus, a hybrid machine learning model (HMLM) is proposed to predict the FE, FF, and PF of disposable facepieces. The HMLM integrates Support Vector Regression (SVR), K-Nearest Neighbors (KNN), and Decision Tree (TREE), with the weights of the three individual models calculated by using the ‌sum of squared errors (SSE)‌. The performance is evaluated using Root Mean Square Error (RMSE). The dataset for training and testing the model is derived from measurement experiments of FE, FF, and PF in a simulated tunnel working environment. The HMLM demonstrated better prediction performance than three individual models for the FE, FF, and PF of disposable facepieces: HMLM's RMSE is 0.0438 (FE), 0.1784 (FF), and 0.2443 (PF); the RMSE of individual models is FE (SVR: 0.0456, KNN: 0.0740, TREE: 0.0941), FF (SVR: 0.2459, KNN: 0.2535, TREE: 0.1864), PF (SVR: 0.4390, KNN: 0.3013, TREE: 0.3265). Compared with the individual model (SVR), HMLM achieves a percentage improvement of 3.9% (FE), 27.5% (FF), and 44.3% (PF), respectively. The findings mitigate the necessity for additional experimental efforts.