基于FIS-CREAM方法的人为差错风险评估

Flight safety assessment based on the human error risk quantification

摘要:

随着航空技术的快速发展，智能化和综合化的机载软硬件设备大大降低了航空事故风险，然而驾驶舱人为差错造成的航空不安全事件并没有得到有效抑制，且人为差错是影响航空安全的重要风险因素之一。本文根据飞机驾驶舱人因特点，结合具体飞行任务，基于模糊推理系统构建飞机驾驶人为差错风险量化模型，以识别飞机操纵过程中关键人为差错并确定人为差错的风险严重性，进而评估航空飞行安全。该方法不仅考虑了人为差错的概率，还考虑了人为差错对驾驶舱系统的影响，并提出了人为差错概率、差错影响概率和人为差错后果作为驾驶舱人为差错风险评估的量化指标。选择进近阶段任务作为研究对象并进行案例分析，实验结果表明，该模型能够精确描述人为差错风险严重性与风险指标之间的关系，降低了专家判断的主观性对结果的影响，且有效解决了数据不足引起的不确定性等问题。本文研究结果能够为减少驾驶舱人为差错，提高飞行安全性且降低航空不安全事件/事故提供有效的理论支持。

Abstract:

With the rapid development of aviation technology, intelligentized and integrated onboard devices highly reduce the risk of aviation accidents. However, aviation unsafety incidents induced by human errors in cockpit are not effectually improved, and human errors are one of the most important risk factors influencing aviation safety. According to the characteristics of human factor in cockpit and a specific flight task, a human error risk quantification model based on the fuzzy inference system is provided to identify key human error types and determine the risk severity of human error. It can further assess flight safety. This method not only considers the probability of human error, but also considers the impact of human error on the cockpit system. This paper employs the human error probability, the error impact probability and the human error consequence to be the risk index to quantify human error risk. Take the approach task as a case, the experimental results show that the model can accurately describe the relationship between risk severity and the three risk indicators, reduce the influence of experts’ subjective judgement on the results and reasonably solve the uncertainty of results caused by insufficient data. This paper can further provide effective theoretical support for the reduction of human errors and the occurrence of aviation accidents.