“…By cumulating probabilities (p,), we obtain the cumulative weight probability distribution from which the weight fraction of the ith pseudocomponent is easily obtained from where the cumulative weight probabilities (Cwis) are estimated by interpolating between the data points with cubic splines. It should be pointed out that an elegant method, the so-called generalized Gaussian quadrature method, developed by Cotterman and Prausnitz (1985) and used elsewhere (Kang and Sandler, 1988;Kang et al, 1989), exists for estimating the pseudocomponents in a mathematically correct and optimal way, given a continuous distribution function for Two sets of criteria are used to determine the optimum number of pseudocomponents. First, the reconstructed EVA M,,, and M,, are calculated according to Mi.…”