Abstract
Abstract The global reliability sensitivity analysis assesses how the model’s failure probability is impacted by uncertainty in input variables, and the importance ranking obtained by the analysis can directly provide effective information on how to decrease the structure’s failure probability. The subset simulation method is combined with the space partition method to solve the variables’ significance measure to overcome the problem of low computational efficiency of traditional methods. In accordance with the failure probability of the subset simulation calculation, the samples generated by this process are statistically processed by the space partition method. Based on the features of stratified sampling in subset simulation, the formula of applicable conditional expectation is derived with the probability stratification information as the weight so as to calculate the random variables’ first-order global reliability sensitivity index. Because the quantity of samples utilized in the proposed method is independent of the dimensions of the problem, the computation’s effectiveness is effectively raised. Ultimately, the suggested approach’s validity is confirmed through a practical example.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call