Cutting planes for the network flow formulation of the single unit commitment problem

Read full text Summarise with VIRAT
Authors: José Ignacio Delgado Anguita
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This thesis work presents an enhanced mixed–integer linear programming (MILP) formulation for the network–flow (NF) based operation of generating units subject to inter–temporal ramping constraints. While ramping limits are essential to ensure feasible transitions between consecutive time periods, their inclusion significantly weakens the linear programming (LP) relaxation of standard formulations, enlarging the fractional feasible region and deteriorating computational performance. To address this issue, a polyhedral study of the ramping polytope was conducted following the scientific method. Initial observations were obtained from a large number of generated instances using the software PORTA, which enabled the formulation of hypotheses regarding potential facet–defining inequalities. These hypotheses were subsequently tested by verifying, through computational experiments, whether the candidate inequalities indeed define facets of the polyhedron and the corresponding conditions. Building upon this analysis, an enhanced NF–based formulation incorporating the derived facet–defining inequalities was proposed. Its performance was evaluated through computational experiments across three different contexts: self-unit commitment (SUC), long-term power system planning, and virtual power plant (VPP) self-scheduling, and compared with state-of-the-art formulations that have been previously shown to be computationally effective. The results demonstrate that the proposed formulation achieves tighter LP relaxations, lower computational effort, and improved scalability with respect to both planning horizon and number of generating units.

Fields of Study & Subjects

Electric Power System Optimization Integrated Energy Systems Optimization Advanced Optimization Algorithms Research

Publication details

Journal / source
CRIS
Publication date
23 March 2026
Language
EN