An improved analytical solution to outcrossing rate for scalar nonstationary and non-gaussian processes

• Published in: Reliability engineering & system safety Vol. 247; p. 110102
• Main Authors: Zhang, Bohua, Wang, Weigang, Lei, Haoran, Hu, Xiancun, Li, Chun-Qing
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2024
• Subjects: Higher-order moments-based reliability index; Non-Gaussian; Nonstationary; Outcrossing rate; Two-component parallel system; Nonstationary; Outcrossing rate; Non-Gaussian; Two-component parallel system; Higher-order moments-based reliability index
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2024.110102

•Propose an analytical method for determining the outcrossing rate of nonstationary and strong non-Gaussian processes.•Combine a two-components parallel system model with higher order moments-based reliability indexes.•This method is more computation efficient than Monte Carlo simulation.•This method is robust for different types of stochastic processes, and for both linear and nonlinear limit state functions. For time-dependent reliability methods, there is a lack of analytical solutions to outcrossing rates of nonstationary and non-Gaussian processes. This paper aims to propose an analytical method to determine this type of outcrossing rate with improved accuracy and efficiency. The novelty of this proposed method is that it can consider the probabilistic properties of non-Gaussian processes by combining a two-component parallel system model with higher-order moments-based reliability indexes. The main contributions of this method are: (1) compared with PHI2+ method, it is more accurate for the calculation of the outcrossing rate of nonstationary and non-Gaussian processes with nonlinear limit state functions; (2) compared with MPHI2 method, it is analytical and not only applicable for the nonstationary stochastic processes but also insensitive to time increments; and (3) it is not only more accurate than the existing methods for nonstationary and non-Gaussian processes but also more computation efficient than the Monte Carlo simulation method. The proposed method has shown its advantages for practical structures with neither Gaussian processes nor linear limit state functions, which are beneficial for both researchers and asset managers to evaluate the time-dependent reliability of structures accurately and to develop risk-informed maintenance schemes with a view to prolonging their service life.