民机维修任务分析的人因可靠性预测模型

Human factor reliability prediction model for civil aircraft maintenance task analysis

人为因素作为人机关系中的重要环节,是影响民航安全的关键因素。定量分析机务人因可靠性水平对于减少无意识犯错、保障飞行安全具有重要意义。根据NASA建议开发出一套THERP+CREAM结合的模型,针对民航特点对THERP数据表进行优化处理,采用模糊贝叶斯网络与CREAM结合的方法确定机务维修人员的认知控制模式,利用平均权重因子将THERP的预测值嵌入CREAM,实现两代方法的结合,以提高机务人因可靠性预测的精度。研究结果表明:通过对某航空公司某维修小组的调查,证实了所提的THERP+CREAM预测模型能够较好地预测机务人因可靠性水平,实现了依据维修手册中的维修任务,即可进行人因可靠性分析,为研制单位维修任务分析中的人素分析提供了一个定量评估方法。

Human factors as an important link in the human-machine relationship are currently becoming a key factor affecting civil aviation safety.Quantitative analysis of human factor reliability level in aircraft operations is important in reducing unconscious errors and ensuring flight safety.In this study,a combined THERP+CREAM model is devel⁃oped according to NASA recommendations,the THERP data table optimized for the characteristics of civil aviation,the cognitive control mode of the aircrew maintenance personnel determined using the combination of fuzzy Bayesian networks and CREAM,and the predicted value of THERP embedded into CREAM using the average weight factor to realize the combination of two generations of methods and improve the accuracy of aircrew human factor reliability pre⁃diction.Investigation into a maintenance team of an airline company shows that the proposed THERP+CREAM predic⁃tion model can better predict the level of airframe human factor reliability,and realize that the human factor reliability analysis can be performed based on the maintenance tasks in the maintenance manual,hence providing a quantitative assessment method for the human factor analysis in the maintenance task analysis of the development unit.