Resumen inglés |
We present a model1ing framework for multistage planning problems under uncertainty in the objective function
coefficients and right-hand-side. A multistagy scenario
analysis scheme with partial recourse is used. So, the decisíon
polícy can be implemented for a given set of initial
time periods (so-called implementable time stage), such
that the solution for the other periods lioes not need' to be
anticipated and, then, it depends upon the scenario group
to occur at each stage. In any ca~e the solution offered for
each stage takes ínto accountal1 scenaríos but wíthout
subordinating to any of them. A novel scheme is presented
for modelling multistage linkíng constraints .through the
scenario tree. This type of constraínts is modelled by using
a splitting variable scheme that allows to produce a sibble
of the coupling variables for each scenario group that belongs
to the latest stage with nonzero coefficients in the
given constraints block. The proposed scheme is very
amenable for usíng decomposition approaches to solve the
deterministic equivalent model and, then, for experimenting
with parallel computing implementations. |