ABSTRACT
ABSTRACT: The application of probabilistic methods for reliability assessment with probabilistic description of random variables in practice is becoming increasingly popular. The new computational methods are still being developed for the calculation of the resulting probability of failure. The paper deals with a new method currently under development: Direct Optimized Probabilistic Calculation (DOProC), which presents a new computational tool for solving probabilistic problems. The method is based on the fundamental principles of probability theory and statistics. Statistically dependent and independent input random quantities (such as load, geometry, material properties, or imperfections) are expressed in the DOProC method by the parametric or empirical distribution prepared in the form of histograms. Calculation of the probability of failure is performed using a specially designed algorithm optimized with numerical integration, which is accurate and efficient for many kinds of probabilistic tasks. This procedure has been programmed in several applications.
