• Corpus ID: 222310722

Learning Active Constraints to Efficiently Solve Bilevel Problems.

  title={Learning Active Constraints to Efficiently Solve Bilevel Problems.},
  author={El'ea Prat and Spyros Chatzivasileiadis},
  journal={arXiv: Optimization and Control},
Bilevel programming can be used to formulate many engineering and economics problems. However, solving such problems is hard, which impedes their implementation in real-life. In this paper, we propose to address this tractability challenge using machine learning classification techniques to learn the active constraints of the lower-level problem, in order to reduce it to those constraints only. Unlike in the commonly used reformulation of bilevel programs with the Karush-Kuhn-Tucker conditions… 

Figures and Tables from this paper

Optimization with Constraint Learning: A Framework and Survey
This paper provides a framework for Optimization with Constraint Learning (OCL) which it is believed will help to formalize and direct the process of learning constraints from data.
Identifying Redundant Constraints for AC OPF: The Challenges of Local Solutions, Relaxation Tightness, and Approximation Inaccuracy
To compute reliable and low-cost operating points, electric power system operators solve optimization problems that enforce inequality constraints such as limits on line flows, voltage magnitudes,
End-to-End Constrained Optimization Learning: A Survey
This paper presents a conceptual review of the recent advancements in integrating combinatorial solvers and optimization methods with machine learning architectures to develop new hybrid machine learning and optimized methods to predict fast, approximate, solutions to combinatorsial problems and to enable structural logical inference.
Cost-driven Screening of Network Constraints for the Unit Commitment Problem
In an attempt to speed up the solution of the unit commitment (UC) problem, both machine-learning and optimization-based methods have been proposed to lighten the full UC formulation by removing as
Offline constraint screening for online mixed-integer optimization
Mixed Integer Linear Programs (MILP) are well known to be NP-hard problems in general, and therefore, tackling and using their solution in online applications is a great challenge. Some Machine


Efficient Evolutionary Algorithm for Single-Objective Bilevel Optimization
This paper introduces bilevel evolutionary algorithm based on quadratic approximations (BLEAQ) of optimal lower level variables with respect to the upper level variables, capable of handling bileVEL problems with different kinds of complexities in relatively smaller number of function evaluations.
Efficiently solving linear bilevel programming problems using off-the-shelf optimization software
Many optimization models in engineering are formulated as bilevel problems. Bilevel optimization problems are mathematical programs where a subset of variables is constrained to be an optimal
Solving Linear Bilevel Problems Using Big-Ms: Not All That Glitters Is Gold
It is shown, through a counterexample, that this widely used trial-and-error approach may lead to highly suboptimal solutions and further research is required to properly select big-M values to solve linear bilevel problems.
An Evolutionary Algorithm for Solving Bilevel Programming Problems Using Duality Conditions
Bilevel programming is characterized by two optimization problems located at different levels, in which the constraint region of the upper level problem is implicitly determined by the lower level
Learning for Constrained Optimization: Identifying Optimal Active Constraint Sets
In many engineered systems, optimization is used for decision making at time scales ranging from real-time operation to long-term planning. This process often involves solving similar optimization
Learning for DC-OPF: Classifying active sets using neural nets
This paper proposes the use of classification algorithms to learn the mapping between the uncertainty realization and the active set of constraints at optimality, thus further enhancing the computational efficiency of the real-time prediction of the optimal power flow.
A new approach for solving linear bilevel problems using genetic algorithms
Technical Note - There's No Free Lunch: On the Hardness of Choosing a Correct Big-M in Bilevel Optimization
It is proved that verifying that a given big-M does not cut off any feasible vertex of the lower level's dual polyhedron cannot be done in polynomial time unless P=NP, and it is shown that verifying this fact is as hard as solving the original bilevel problem.
Bilevel programming approach to demand response management with day-ahead tariff
  • A. Kovács
  • Computer Science
    Journal of Modern Power Systems and Clean Energy
  • 2019
The primal-dual reformulation is proposed in this paper to convert the bilevel optimization problem into a single-level quadratically constrained quadratic program (QCQP), and a successive linear programming (SLP) algorithm is applied to solve it.
Statistical Learning for DC Optimal Power Flow
An ensemble control policy is proposed that combines several basis policies to improve performance and is based on the observation that the OPF solution corresponding to a certain uncertainty realization is a basic feasible solution, which provides an affine control policy.