摘要
综述了为提高失效概率计算效率的研究成果,包括基于抽样的高效失效概率算法和基于积分的高效失效概率算法。其中,基于抽样的高效失效概率算法在传统蒙特卡洛模拟方法的基础上,通过重要性抽样方法在失效域抽样、最优抽样技术优化分区样本量、分区细化技术减少分区数量,从而减少蒙特卡洛模拟样本量。另外,基于积分的高效失效概率算法通过建立N次飞行循环与初始循环(N=0)随机变量空间的映射关系,解决了时变失效区域中概率密度函数难以求解的困难。在与蒙特卡洛相对误差小于5%条件下,积分算法时间成本降低了数十倍。
The research results for improving the efficiency of failure probability calculation,including the efficient sampling-based algorithms and the efficient integration-based algorithms, were summarized. Among them, based on the traditional Monte Carlo simulation method, the importance sampling algorithm generated samples in the failure domain. The optimal sampling technology optimized the zone sample size. The zone refinement technology reduced the number of zones,thus reducing the Monte Carlo simulation sample size. In addition,the direct integration was realized by establishing the mapping relationship of failure domain at N flight cycles and the initial(N = 0) flight cycles based on the probability density theory. When the relative error with Monte Carlo was less than 5%,the calculation time cost was reduced by at least tens of times.
作者
李果
刘俊博
周惠敏
丁水汀
LI Guo;LIU Junbo;ZHOU Huimin;DING Shuiting(School of Energy and Power Engineering,Beihang University,Beijing 100191,China;Civil Aviation University of China,Tianjin 300300,China)
出处
《航空动力学报》
EI
CAS
CSCD
北大核心
2022年第11期2398-2407,共10页
Journal of Aerospace Power
关键词
航空发动机
寿命限制件
高能转子轮盘
概率失效风险评估
高效失效概率算法
数值积分算法
蒙特卡洛方法
aero-engine
life limited parts
high energy rotor disk
probabilistic failure risk assessment
efficient algorithm of failure probability
numerical integration algorithm
Monte Carlo method