This paper, based on a plant-specific, uses Bayesian model to predict the effects of shift workers' operation on industrial risk in chemical process. Chemical Process Industries (CPI) security risk mainly originates from the aging of physical devices and erroneous human performance. Furthermore, the operational skills of shift workers and interaction between different teams most likely affect the reliability of the chemical processes. Gaussian copulas can reflect the complex nonlinear relationship between variables, therefore, the failure probabilities of the safety system and the shift workers are estimated by using Gaussian copulas and Bayesian analysis to ensure better predictions. Herein, considering the fact that the operational sequences of shift workers and interaction between different teams tend to affect chemical plants risk, the characteristics of the time series and coupling in various teams' operation are proposed. Gaussian copula shows the relationship between operators and safety control systems and the interaction of shift workers, then using Bayesian model to quantitatively evaluate the risk, and the results can be more suitable to the engineering facts.%基于贝叶斯模型以特定化工生产装置为例,定量分析班组人员操作对化工风险的影响.化工生产安全风险主要来自于设备的老化和人员的误操作.轮班班组的操作能力以及不同班组之间的关联性很可能影响化工过程的可靠性.Gaussian copula可以体现变量之间复杂的非线性关系,因此利用贝叶斯分析结合Gaussian copula可以较好地评估安全系统和轮班班组风险.在此,提出了不同班组操作的时序性和耦合性来体现班组操作的顺序性和班组间的关联性等特征.Gaussian copula函数体现班组与安全控制系统之间以及班组之间的关联性,进而利用贝叶斯模型进行定量风险评估,结果更切合实际.
展开▼
