Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

GO Competition Challenge 3: Problem, Solvers, and Solution Analysis

Jesse T. Holzer, Stephen Elbert, Hans Mittelmann, Richard O'Neill, HyungSeon Oh

Abstract

This paper describes the Grid Optimization (GO) Competition Challenge 3, focusing on the problem motivation, formulation, solvers submitted by competition entrants, and analysis of the solutions produced. Funded by DOE/ARPA-E and led by a collaboration of national labs and academia members, the GO Competition addresses challenging problems in power systems planning and operations to drive research in advanced solution methods essential for a rapidly evolving electric power sector. Challenge 3 targets a multi-period unit commitment problem, incorporating AC power modeling and topology switching to reflect the dynamic grid management techniques required for future power systems. The competition results offer significant benefits to both researchers and industry practitioners. For researchers, it fosters innovation, encouraging the development of new algorithms to address the complexities of modern power systems. For industry practitioners, the competition drives the creation of more efficient and reliable computational tools, directly improving grid management practices. This collaboration bridges the gap between theory and practical implementation, advancing the field in meaningful ways. This paper documents the problem formulation, solver approaches, and the effectiveness of the solutions developed.

GO Competition Challenge 3: Problem, Solvers, and Solution Analysis

Abstract

This paper describes the Grid Optimization (GO) Competition Challenge 3, focusing on the problem motivation, formulation, solvers submitted by competition entrants, and analysis of the solutions produced. Funded by DOE/ARPA-E and led by a collaboration of national labs and academia members, the GO Competition addresses challenging problems in power systems planning and operations to drive research in advanced solution methods essential for a rapidly evolving electric power sector. Challenge 3 targets a multi-period unit commitment problem, incorporating AC power modeling and topology switching to reflect the dynamic grid management techniques required for future power systems. The competition results offer significant benefits to both researchers and industry practitioners. For researchers, it fosters innovation, encouraging the development of new algorithms to address the complexities of modern power systems. For industry practitioners, the competition drives the creation of more efficient and reliable computational tools, directly improving grid management practices. This collaboration bridges the gap between theory and practical implementation, advancing the field in meaningful ways. This paper documents the problem formulation, solver approaches, and the effectiveness of the solutions developed.

Paper Structure

This paper contains 41 sections, 39 equations, 13 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Importance, Challenges, and Implications of Solving the GO Competition Challenge 3 Problem of Multi-Period Unit Commitment with AC Power Flow
  3. Problem Description
  4. Literature Survey
  5. Problem Formulation
  6. Nomenclature
  7. Index sets
  8. Distinguished set elements
  9. Subsets
  10. Parameters
  11. Functions
  12. Decision variables
  13. Sets and indices
  14. State variables and complex coordinates
  15. Device settings and feasible regions
  16. ...and 26 more sections

Figures (13)

  • Figure 1: System-wide aggregate real power supply and demand curves for a single time interval in a 73-bus problem instance.
  • Figure 2: Real power imbalance penalty as a percent of total objective vs. total objective in division 1 ensemble solutions.
  • Figure 3: Reactive power imbalance penalty as a percent of total objective vs. total objective in division 1 ensemble solutions.
  • Figure 4: Base case line overload penalty as a percent of total objective vs. total objective in division 1 ensemble solutions.
  • Figure 5: Reserve imbalance penalty as a percent of total objective vs. total objective in division 1 ensemble solutions.
  • ...and 8 more figures