论文部分内容阅读
以结冰条件下的飞行风险量化概率为研究对象,基于蒙特卡罗飞行仿真实验对结冰条件下人-机-环系统的耦合特性进行了分析,并获取了飞行参数极值样本。构建了飞行风险发生的判定条件;对飞行参数极值样本进行了统计特性分析,验证了其厚尾分布特征。一维分布类型辨识结果表明广义极值分布对相对速度和迎角极值的描述精度最高。为描述二维变量对相关性的各自影响程度,提出了一种新的双参数变权重Copula模型;辨识结果表明该Copula模型能以较高的精度通过假设检验。相关性分析的结果表明相对速度和迎角同时出现极大值和极小值的概率较大。基于二维极值样本的Copula分布模型求出了不同结冰程度下的飞行风险概率值,探讨了飞行风险的非线性增长趋势。
Taking the quantification probability of flight risk under icing condition as the research object, the coupling characteristics of the man-machine-loop system under icing condition were analyzed based on the Monte Carlo flight simulation experiment and the extreme value samples of flight parameters were obtained. The judgment conditions of the occurrence of flight risk are constructed. The statistical characteristics of the extreme value flight parameters are analyzed, and the characteristics of their fat tail distribution are verified. The results of one-dimensional distribution type identification show that the generalized extreme value distribution has the highest accuracy of describing the relative speed and the angle of attack. To describe the influence of two-dimensional variables on the correlation, a new two-parameter variable weight Copula model is proposed. The recognition results show that the Copula model can pass the hypothesis test with high accuracy. The result of the correlation analysis shows that there is a high probability that both the relative speed and the angle of attack have both the maximum value and the minimum value. Based on the Copula distribution model of two-dimensional extreme value samples, the probability of flight risk under different degrees of freezing was calculated and the non-linear trend of flight risk was discussed.