A Multilevel Simulation Method for Time-Variant Reliability Analysis

Crude Monte Carlo simulation (MCS) is the most robust and easily implemented method for performing time-variant reliability analysis (TRA). However, it inefficient, especially high problems. This paper aims to present a random called multilevel (MLMC) TRA enhance computational efficiency of crude MCS while maintaining its accuracy robustness. The proposed first discretizes time interval interest using geometric sequence different timesteps. cumulative probability failure associated with finest level can then be estimated by computing corrections all levels. To assess in way that minimizes overall complexity, number samples at each optimized. Moreover, correction independently computed MCS. Thereby, achieve much lower cost than MCS, retains robustness respect nonlinearity dimensions. effectiveness validated numerical examples.