Let be the expected performance measure of a discrete event system (DES), where L is the sample performance depending on the vector of parameters ν2 and driven by an input vector Y, which has a probability density function (pdf) is a vector of parameters, and the subscript ν1 in indicates that the expectation is taken with respect to the pdf . Suppose that is not available analytically and it is required to evaluate (estimate) it, as well as the associated sensitivities simultaneously for different values of via simulation. This paper shows that in some cases interesting for applications, and can be estimated by using the so-called ‘push out’ technique. More precisely, it is shown that it is possible to replace the original sample performance by an auxiliary one while ‘pushing out’ the parameter vector from the original sample performance function to a pdf associated with the original one . The paper also shows how both the auxiliary sample performance and the associated pdf can be obtained from their original counterparts and how to combine them together to perform sensitivity analysis for the original DES. Particular emphasis will be placed on the case where the sample performance function is neither analytically available nor everywhere differentiable in. The paper finally discusses the advantage of the proposed method and present numerical results supporting the presents theory.
