数据驱动的操纵员情景意识因果模型研究

为克服传统情景意识(SA)可靠性评价方法的不足,建立了更具鲁棒性的SA因果模型。首先,采用组织定向的SA失误分析框架或方法对核电厂人因事件进行分析,获得了132组样本数据。然后,采用相关性分析方法识别SA影响因素的相关关系,并采用因子分析方法识别SA失误发生的场景,包括操纵员的心智水平、工作态度、压力水平及系统状态呈现水平等。最后,基于上述研究结果,建立数据驱动的SA因果模型。结果表明,基于数据驱动的SA因果模型识别了SA失误发生的场景,且考虑了行为形成因子的因果关系有利于提升SA可靠性定量评价的精度。

Study on Data-driven Operator’s Situational Awareness Causality Model

In order to overcome the shortcomings of traditional assessment method of situational awareness (SA) reliability, a more robust SA causality model was built in this paper. Firstly, the organization-oriented SA error analysis framework or method was adopted to analyze nuclear power plant incidents and 132 samples were obtained. Then, the correlation analysis method was used to identify the correlation relationships between factors of influencing SA and factor analysis method was used to identify the scenes triggering SA errors, including operator’s mental level, operator’s work attitude, stress level and system situation display level. Finally, based on the above analysis results, a data-driven SA causality model was established. The results show that the data-based SA causality model can identify the scenes triggering SA errors. It is useful to improve the accuracy of quantitative assessment of SA reliability because of considering the causality relationships of performance shaping factors.