论文部分内容阅读
提出了结合极值理论与Copula模型来量化评估平尾结冰条件下飞行风险概率的方法。通过建立人-机-环复杂系统模型,对平尾在进近与着陆过程中的结冰情形进行仿真,采用蒙特卡罗法提取平尾结冰极值参数,验证了所提取极值参数符合一维广义极值(GEV)分布,根据飞行风险的定义和相关安全性准则,建立了平尾结冰飞行风险发生的判定条件,计算得出一维极值飞行风险概率;在此基础上选取Copula模型来描述二维极值参数的相关性,对多种Copula模型的未知参数进行辨识,通过拟合优度检验对精度进行验证,得出Joe Copula模型对二维极值分布的描述最为准确,运用Joe Copula模型计算出二维极值飞行风险概率,有效解决了一维极值具有的局限性。所提方法对飞行安全评估等理论有一定参考价值,能为平尾结冰飞行事故的预防提供分析和检验依据。
A method combining the extreme value theory and the Copula model to quantitatively evaluate the flight risk probability under the flat-tail icing condition is proposed. Through the establishment of the complex system model of human-machine-loop, the icing conditions of the tail during approach and landing are simulated, and the ice tail extreme ice parameters are extracted by Monte-Carlo method. It is verified that the extracted extreme value parameters conform to the one-dimensional Generalized Extreme Value (GEV) distribution. According to the definition of flight risk and the related safety criterion, the judgment conditions of the flight-end icing flight risk are established and the probability of one-dimensional extreme flight risk is calculated. Based on this, the Copula model Describe the correlation of two-dimensional extreme parameters, identify unknown parameters of a variety of Copula models, verify the accuracy by goodness-of-fit test, draw the conclusion that Joe Copula model has the most accurate description of two-dimensional extreme distribution, Copula model to calculate the probability of two-dimensional extreme flight risk, effectively solve the one-dimensional limit has its limitations. The proposed method has certain reference value to the theory of flight safety assessment, which can provide the basis for analysis and test for the prevention of Pingtai icing flight accident.