Evaluation of the probability density functions of distribution system reliability indices with a characteristic functions-based approach

In reliability analysis of distribution systems, random events like the occurrence of a fault or the time to restore the service after a fault are represented by using random variables (RVs), so that the reliability indices built on the basis of these RVs also become RVs. Existing techniques for the evaluation of the probability distributions of reliability indices are typically based on Monte Carlo and analytical simulations. This paper presents a new method for computing the probability distribution of reliability indices. The random sums introduced by the randomness of the number of fault occurrences in the time interval of analysis are handled by using a characteristic functions-based approach. The direct convolution of the probability density functions is avoided by resorting to the properties of the compound Poisson process. In addition, the direct and inverse discrete Fourier transforms are used to allow for handling any type of probability distribution. The proposed method is an effective alternative to the existing methods, providing a fast and simple computation of probability distributions and moments for local and global reliability indices. Results obtained for large real urban distribution systems are presented.