基于FTA-ANN的航空发动机安全风险分析

Safety Risk Analysis of Aero-engine Based on FTA-ANN

航空发动机属于多发性故障机械,是典型的高风险配件,航空公司对航空发动机安全都给予了高度重视,对航空发动机可靠性进行评估,对发动机发生的危害性后果进行深入分析,对于准确评估航空发动机的安全风险水平,提出针对性的管理策略具有重要意义。故提出面向航空发动机,利用故障树映射到ANN的方法,实现动态安全风险分析。以某型飞机某型发动机危害性后果“发动机起火不可控”为案例,对该危险后果进行故障树分析,对其建立故障树并一一映射,采用人工神经网络ANN进行多次训练。利用算例,将得到结果与FT结果进行比较发现:ANN与FT结果匹配率较高,说明以FT映射的ANN模型是一种有效的风险评估技术。所提出的方法还可以应用到核电站系统、电力系统等其他高可靠性的复杂系统中。

Aeroengines are multi-fault machines and typical high-risk parts.Airlines pay great attention to aero-engine safe⁃ty,evaluate aero-engine reliability,and conduct in-depth analysis of the harmful consequences of engines.It is of great signifi⁃cance to propose a targeted management strategy for the safety risk level of the engine.Therefore,it is proposed to target the aeroen⁃gine and use the fault tree to map to the ANN to realize dynamic security risk analysis.Taking the dangerous consequences of a cer⁃tain type of engine"uncontrollable engine fire"as an example,the fault tree is analyzed for the dangerous consequences,and the fault tree is built and mapped one by one.The artificial neural network ANN is used for multiple trainings.Using the example,the results are compared with the FT results.It is found that the matching rate between ANN and FT results is high,indicating that the ANN model with FT mapping is an effective risk assessment technique.The proposed method can also be applied to other high-reli⁃ability complex systems such as nuclear power plant systems and power systems.