Probabilistic load flow using multilinearisations

Probabilistic load flow is a technique which recognises the probabilistic nature of the input parameters and takes account of the uncertainties and random variations of generation and load within one computer run. Many formulations have been presented to implement this technique. The majority have assumed either, a linear model to represent the load flow equations, or the output random variables to be normally distributed, or a combination of both of these concepts. This paper, first, uses Monte Carlo simulation to show the effects of nonlinearity in the network equations and that the assumption of a normal distribution for the output random variables is completely unreliable. A completely new algorithm for the probabilistic load flow is then proposed which accounts for the nonlinear effects. This new technique is shown to give a new impetus to probabilistic load flow analysis