A simplified but accurate approach for reliability computation of a complex system

A methodology for the reliability computation of complex systems is presented. It is assumed that the individual s-independent random variables composing the strength and stress of a given system are s-normal. The functional relationship is usually complex. As a result, the derivation of the distribution of stress or strength exceeds the capacity of common transformation or convolution techniques. The discretizing approach recommended by J.R. D´Errico and N.A. Zaino (1988) significantly eases the computational burden in the determination of stress or strength distributions. This is achieved by discretizing each of the random components composing the stress or strength of a system. As a result, probability mass functions are utilized to estimate the strength and stress of the system. This methodology presents an opportunity for significant improvement in the computational efficiency for reliability computation of a complex system