Compact trees: a new approach to model uncertainty for the management of power systems

For the management of power systems under market conditions, dealing with uncertainties is a very important task. But when several uncertainties are to be represented in a scenarios tree, the tree often becomes too big, making the calculation impossible. This paper presents a method to combine a random process with a given memory (representing the load) and another one, non time-correlated (generation units availability), in order to enable the resolution by an algorithm using the dynamic programming principle without generating too huge a tree.