In reliability-based design (RBD), uncertainties are usually treated stochastically, and nondeterministic variables are assumed to follow certain probability distributions. However, in many practical engineering applications, distributions of some random variables may not be precisely known or uncertainties may not be appropriately represented with distributions. The possible values of those nondeterministic variables are often only known to lie within specified intervals without precise distribution information. In this paper, we attempt to address this issue by proposing a RBD method to deal with the uncertain variables characterized by the mixture of probability distributions and intervals. The reliability is considered under the condition of the worst case combination of interval variables. The computational demand of RBD with the mixture of random and interval variables may increase dramatically due to the need for identifying the worst case interval variables. To alleviate the computational burden, a sequential single-loop procedure is employed to replace the computationally expensive double-loop procedure when the worst case scenario is applied directly. with the proposed method, the RBD is conducted within a series of cycles of deterministic optimization and reliability analysis. The optimization model in each cycle is built based on the most probable point under the worst case combination of the interval variables obtained from the reliability analysis in the previous cycle. Since the optimization is decoupled from the probabilistic analysis, the computational amount for reliability analysis is decreased to the minimum extent. The proposed method is demonstrated with two examples.
National Science Foundation (U.S.)
- Risk Management,
- Design Engineering,
- Random Processes,
Available at: http://works.bepress.com/xiaoping-du/64/