 |
Enhanced Cutting-Plane Algorithm for Stochastic Unit Commitment Using Interval Variables
1 : École nationale des ponts et chaussées
Institut Polytechnique de Paris
2 : EDF R&D
EDF
7 boulevard Gaspard Monge, 91120 Palaiseau -
France
The Unit Commitment (UC) problem is a fundamental challenge for energy producers and network operators aiming to determine an optimal generation schedule for electricity production units, such as thermal and hydro units, while adhering to technical and economic constraints.
A major source of complexity in UC arises from uncertainty, particularly in forecasting energy demand and wind power generation. To address stochastic UC, cutting-plane algorithms have been proposed in the literature to handle large numbers of scenarios efficiently
Recently, a new formulation of the deterministic UC problem based on interval variables has been proposed to better model unit production decisions. In this talk, we show that, to solve stochastic UC, incorporating these interval variables when commitment decisions are made in the first stage enables the derivation of stronger cuts, leading to a significant improvement in the convergence of the algorithm.
Through computational experiments, we demonstrate that our method significantly improves solution efficiency, enabling the resolution of large-scale instances. In particular, we show that our approach performs efficiently under both a risk-neutral framework and an average value-at-risk measure.
