Kaplanvural Selcan, Tosyalı Eren, Ekmekçi İsmail
Vocational School of Health Services, Occupational Health and Safety, İstanbul Bilgi University, Istanbul, 34387, Turkey.
Department of Occupational Health and Safety, Institute of Science, Istanbul Commerce University, 34840, Istanbul, Turkey.
Sci Rep. 2024 Oct 14;14(1):23985. doi: 10.1038/s41598-024-74938-9.
The motivation of this study is to develop and establish an occupational accident dynamical model (OA model) based on Susceptible-Infected-Recovered framework. In order to investigate the dynamics of the OA model, monthly occupational accident data from Turkey between 2013 and 2020 has been selected as dataset. The OA model is defined by a coupled first-order ordinary nonlinear differential equation with four variables. In addition, the relationships between these variables are described with ten parameters. The OA model's characterization of the equilibrium points is analyzed by investigating the behaviors of these points according to the eigenvalues derived from the Jacobian matrix. Also, the stability of these points is obtained according to the eigenvalues. These results show the behavior of the system near equilibrium points. After that, the reproduction number is computed by using the next-generation matrix method. The calculated reproduction number for given parameters reveals that the OA model is unstable. The OA model is numerically solved in 96 steps, with a time interval of 1 month, using the ODE45 Matlab routine based on the explicit Runge-Kutta algorithm. In addition, a modified OA model is developed by adding the Occupational Health and Safety (OHS) re-training parameter to the OA model to observe a reducing effect on occupational accident numbers. The main results of this study provide a new approach about the future estimation of the number of occupational accidents. Furthermore, through the comparison of numerical results from both models, the study demonstrates that national safety policies, particularly those enhancing the efficacy of OHS training, can effectively mitigate accidents.
本研究的目的是基于易感-感染-康复框架开发并建立一个职业事故动态模型（OA模型）。为了研究OA模型的动态特性，选取了2013年至2020年土耳其的月度职业事故数据作为数据集。OA模型由一个具有四个变量的耦合一阶常非线性微分方程定义。此外，这些变量之间的关系用十个参数来描述。通过根据从雅可比矩阵导出的特征值研究这些平衡点的行为，分析了OA模型平衡点的特征。同时，根据特征值得到这些点的稳定性。这些结果显示了系统在平衡点附近的行为。之后，使用下一代矩阵方法计算繁殖数。给定参数下计算出的繁殖数表明OA模型是不稳定的。使用基于显式龙格-库塔算法的ODE45 Matlab例程，以1个月的时间间隔，分96步对OA模型进行数值求解。此外，通过在OA模型中添加职业健康与安全（OHS）再培训参数，开发了一个改进的OA模型，以观察其对职业事故数量的减少效果。本研究的主要结果为职业事故数量的未来估计提供了一种新方法。此外，通过比较两个模型的数值结果，该研究表明国家安全政策，特别是那些提高OHS培训效果的政策，可以有效减少事故。
