A bilevel framework for decision-making under uncertainty with contextual information

Muñóz, Miguel Ángel Suárez; Pineda, Salvador

doi:10.1016/j.omega.2021.102575

Cited by 18 publications

(14 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, in Muñoz et al (2022) a bi-level framework is proposed to fit a parametric model to those data that are specifically tailored to maximize the decision value, while accounting for possible feasibility constraints. In Bertsimas et al (2019) and Esteban-Pérez and Morales (2021), parametric approaches are left behind to focus on estimating a complete conditional distribution of the side information to make robust decisions over x.…”

Section: Introductionmentioning

confidence: 99%

Optimal day-ahead offering strategy for large producers based on market price response learning

Alcántara¹,

Ruiz²

2022

Preprint

View full text Add to dashboard Cite

In day-ahead electricity markets based on uniform marginal pricing, small variations in the offering and bidding curves may substantially modify the resulting market outcomes. In this work, we deal with the problem of finding the optimal offering curve for a risk-averse profit-maximizing generating company (GENCO) in a data-driven context. In particular, a large GENCO's market share may imply that her offering strategy can alter the marginal price formation, which can be used to increase profit. We tackle this problem from a novel perspective. First, we propose a optimization-based methodology to summarize each GENCO's step-wise supply curves into a subset of representative price-energy blocks. Then, the relationship between the market price and the resulting energy block offering prices is modeled through a Bayesian linear regression approach, which also allows us to generate stochastic scenarios for the sensibility of the market towards the GENCO strategy, represented by the regression coefficient probabilistic distributions. Finally, this predictive model is embedded in the stochastic optimization model by employing a constraint learning approach. Results show how allowing the GENCO to deviate from her true marginal costs renders significant changes in her profits and the market marginal price. Furthermore, these results have also been tested in an out-of-sample validation setting, showing how this optimal offering strategy is also effective in a real-world market contest.

show abstract

Section: Introductionmentioning

confidence: 99%

Optimal day-ahead offering strategy for large producers based on market price response learning

Alcántara¹,

Ruiz²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Within this context, two possible avenues of research are opened to achieve better results: i) a focus on improving the decision-making model (prescriptive framework), which assumes we can change it to consider embedded co-optimized forecasts [146]; or ii) a focus on improving the forecasting model (predictive framework), which assumes we can not change the decision-making process (in our application, defined by system operators' dispatch models), but we can change the forecasts to incorporate, in a closed-loop manner, a given application cost function [59,151,152]. Therefore, in this paper, we focus on the latter avenue.…”

Section: ¦ ¥mentioning

confidence: 99%

“…Additionally, our scalable heuristic method ensures optimal second-level solutions. This is a salient feature of our method in contrast with other methods that rely on surrogates of the second level [151] or solve approximations of the KKT conditions with non-linear solvers [152]. In this context, the proposed framework is general and suitable for a wide range of applications relying on the standard structure of the forecastdecision process.…”

Section: Objective and Contributionmentioning

confidence: 99%

“…While working on this paper, the authors became aware of the work by [152] (developed simultaneously with ours, by a completely different team) that guards many conceptual similarities to the general version of our proposed model. The work of [152] also focuses on the idea of finding the best forecast for a given application (or context) through a bilevel framework. The framework proposed in [152] is applied to estimate a parameter of the inverse demand curve of a Cournot strategic producer bidding in forward markets.…”

Section: Application-driven Forecast Modelsmentioning

confidence: 99%

“…The work of [152] also focuses on the idea of finding the best forecast for a given application (or context) through a bilevel framework. The framework proposed in [152] is applied to estimate a parameter of the inverse demand curve of a Cournot strategic producer bidding in forward markets. The scalability of their model relies on a nonlinear relaxation of the righthand side of the complementarity constraints.…”

Section: Application-driven Forecast Modelsmentioning

confidence: 99%

See 2 more Smart Citations

The Effectiveness of Bilevel Optimization in Large-Scale Power Systems Problems: A Bilevel Optimization Toolbox, a Framework for Application-Driven Learning, and a Market Simulator

GARCIA¹

View full text Add to dashboard Cite

First and foremost, I would like to thank Beatriz for all the companionship and support of my friend, partner, girlfriend, and wife, Beatriz, who helped me with absolutely everything during my doctoral years. Your support, affection, attention, help, and lots of love made this journey much more peaceful and pleasant.I would like to thank my family for their constant support, motivation, excitement, interest, and much love. My parents, Patricia and Joaquim, gave me every opportunity and always encouraged me. My grandparents are all examples that contributed a lot to my path. Helena always motivates my curiosity and willingness to explore everything. Maria Olivia always brings the whole family together to give more strength to everyone. Carlos José, the most creative person I know always excited about my academic developments. Luiz Antônio, is an incredible example of kindness and goodness, unfortunately, he could not see the end of this journey here on earth for a few months, but he has certainly seen it from wherever he is. My brother João Pedro who was always by my side at all times, always ready to do anything for me. Luiz Gabriel and Rafael who are my hope for the future of this world. Maria Teresa and Steve who have been so supportive over the years. My cousins Maria, Luiza and Carol for their company and support.Thanks to my dear friends who have always been by my side on this journey. Especially when I needed to work on this thesis.I would like to thank my advisor and great friend Alexandre Street. Thank you for your company and guidance on this long journey that began as a scientific initiation project in the electrical engineering graduation. Since then, we've worked hard, and I've learned how to write, present, research, converse, and much more. He was a sensational guide on this trip, and I could not have chosen someone better to develop this work with.I thank Mario Veiga, my advisor, supervisor, friend, and one of my heroes. Mario shows every day how to do top-quality science. Since my first day of internship at PSR, Mario has taught me to be a better researcher, developer, professional, and person. Our daily discussions motivated the works of this thesis, among many others. Thank you for introducing me to JuMP, which was essential throughout the thesis.Thanks to my incredible colleagues at PUC and LAMPS. Special thanks to

show abstract

Deep Learning Based Bidding-Oriented Probability Density Forecasting Approach for Renewable Energy Generation in Virtual Power Plant

Wang,

et al. 2024

Environmental Science and Engineering

View full text Add to dashboard Cite

A bilevel framework for decision-making under uncertainty with contextual information

Cited by 18 publications

References 25 publications

Optimal day-ahead offering strategy for large producers based on market price response learning

Optimal day-ahead offering strategy for large producers based on market price response learning

The Effectiveness of Bilevel Optimization in Large-Scale Power Systems Problems: A Bilevel Optimization Toolbox, a Framework for Application-Driven Learning, and a Market Simulator

Deep Learning Based Bidding-Oriented Probability Density Forecasting Approach for Renewable Energy Generation in Virtual Power Plant

Contact Info

Product

Resources

About