A New Metric for Time-Dependent Mechanism Reliability Analysis

In classical time-dependent mechanism reliability analysis, probability or fuzzy reliability theory is employed. However, the method based on probability theory requires probability distributions of all the uncertain parameters, but the accurate distributions of parameters in real life are sometimes difficult to obtain. Moreover, fuzzy theory fails to describe many subjective random variables due to the duality of randomness. In this paper, we utilize uncertain variables to uniformly represent the subjective random and epistemic uncertainties by quoting the uncertainty theory. A new quantification method based on the uncertain variables is presented in order to simultaneously satisfy the duality of subjective random and subadditivity of epistemic uncertainties. In addition, first order Taylor series expansion is used to deal with nonlinear limit state functions. A novel mechanism uncertainty reliability metric (URM) is proposed based on normal uncertainty distribution. Besides, point kinematic reliability analysis method combining with uncertainty theory is presented to estimate the URM of the mechanism at each time instant. Finally, by applying the proposed method to a practical engineering example, the trend of uncertainty reliability is consistent with the traditional method. The results show that the traditional method ignores the epistemic uncertainty will overestimate the reliability of the time-dependent mechanism. The proposed method in this paper can provide a better assessment for the time-dependent mechanism reliability analysis under epistemic uncertainties.