multistage stochastic programming

Participation in the 31st European Conference on Operational Research.

Last July 21th, the work "Multistage Scenario Trees Generation for Electricity Markets Optimization" was presentated at the 31st European Conference on Operational Research  in an invited session of the stream "Stochastic and Robust Optimization". This study is a result of the PH.D. Thesis of Ms. Marlyn D. Cuadrado on scenarios trees generation for multistage stochastic programming in optimal electricity bid problems.

Multistage Scenario Trees Generation for Electricity Markets Optimization

Publication TypeConference Paper
Year of Publication2021
AuthorsMarlyn Dayana Cuadrado Guevara; F.-Javier Heredia
Conference Name31st European Conference on Operational Research.
Conference Date11-14/07/2021
Conference LocationAthens
Type of WorkInvited presentation
ISBN NumberISBN 978-618-85079-1-3
Key Wordsresearch; multistage stochastich programming; virtual power plants; electricity markets; scenarios tree generation
AbstractThe presence of renewables in electricity markets optimization have generated a high level of uncertainty in the data, which has led to a need for applying stochastic optimization to model this kind of problems. In this work, we apply Multistage Stochastic Programming (MSP) using scenario trees to represent energy prices and wind power generation. We developed a methodology of two phases where, in the first phase, a procedure to predict the next day for each random parameter of the MSP models is used, and, in the second phase, a set of scenario trees are built through Forward Tree Construction Algorithm (FTCA) and a modified Dynamic Tree Generation with a Flexible Bushiness Algorithm (DTGFBA). This methodology was used to generate scenario trees for the Multistage Stochastic Wind Battery Virtual Power Plant model (MSWBVPP model), which were based on MIBEL prices and wind power generation of a real wind farm in Spain. In addition, we solved three di erent case studies corresponding to three di erent hypotheses on the virtual power plant’s participation in electricity markets. Finally, we study the relative performance of the FTCA and DTGFBA scenario trees, analysing the value of the stochastic solution through the Forecasted Value of the Stochastic Solution (FVSS) and the classical VSS for the 366 daily instances of the MSWBVPP problem spanning a complete year.
URLClick Here
ExportTagged XML BibTex

Ph D. Thesis on multistage scenario tree generation for renewable energies.

 On November 30th 2020 took place the defense of the Ph.D. Thesis entittled "Multistage Scenario Trees Generation for Renewable Energy Systems Optimization", authored by Ms. Marlyn D. Cuadrado Guevara and advised by prof. F.-Javier Heredia. In this thesis a new methodology to generate and validate probability scenario trees for multistage stochastic programming problems arising in two different energy systems with renewables are proposed. The first problem corresponds to the optimal bid to electricity markets of a virtual power plant that consists on a wind-power plant plus a battery storage energy systems. The second one is the optimal operation of a distribution grid with some photovoltaic production.

Multistage Scenario Trees Generation for Renewable Energy Systems Optimization

Publication TypeThesis
Year of Publication2020
AuthorsMarlyn Dayana Cuadrado Guevara
Academic DepartmentDept. of Statistics and Operations Research. Prof. F.-Javier Heredia, advisor.
Number of Pages194
UniversityUniversitat Politècnica de Catalunya-BarcelonaTech
CityBarcelona
DegreePhD Thesis
Key Wordsresearch; Battery energy storage systems; Electricity markets; Ancillary services market; Wind power generation; Virtual power plants; Multistage Stochastic programming; phd thesis
AbstractThe presence of renewables in energy systems optimization have generated a high level of uncertainty in the data, which has led to a need for applying stochastic optimization to modelling problems with this characteristic. The method followed in this thesis is Multistage Stochastic Programming (MSP). Central to MSP is the idea of representing uncertainty (which, in this case, is modelled with a stochastic process) using scenario trees. In this thesis, we developed a methodology that starts with available historical data; generates a set of scenarios for each random variable of the MSP model; defines individual scenarios that are used to build the initial stochastic process (as a fan or an initial scenario tree); and builds the final scenario trees that are the approximation of the stochastic process.
URLClick Here
ExportTagged XML BibTex

A multistage stochastic programming model for the optimal bid of a wind producer

Publication TypeConference Paper
Year of Publication2018
AuthorsF.-Javier Heredia; Marlyn D. Cuadrado; J.-Anton Sánchez
Conference Name23th International Symposium on Mathematical Programming
Conference Date01-06/07/2018
Conference LocationBordeaux
Type of Workcontributed presentation
Key Wordsresearch; Battery energy storage systems; Electricity markets; Ancillary services market; Wind power generation; Virtual power plants; Stochastic programming
AbstractAbstract: Battery Energy Storage Systems (BESS) can be used by wind producers to improve the operation of wind power plants (WPP) in electricity markets. Associating a wind power plant with a BESS (the so-called Virtual Power Plant (VPP)) provides utilities with a tool that converts uncertain wind power production into a dispatchable technology that can operate not only in spot and adjustment markets (day-ahead and intraday markets) but also in ancillary services markets that, up to now, were forbidden to non-dispatchable technologies. We present in this study a multi-stage stochastic programming model to find the optimal operation of a VPP in the day-ahead, intraday and secondary reserve markets while taking into account uncertainty in wind power generation and clearing prices (day-ahead, secondary reserve, intraday markets and system imbalances). A new forecasting procedure for the random variables involved in stochastic programming model has been developed. The forecasting model is based on Time Factor Series Analysis (TFSA) and gives suitable results while reducing the dimensionality of the forecasting mode. The quality of the scenario trees generated using the TFSA forecasting models with real electricity markets and wind production data has been analysed through multistage VSS.
URLClick Here
ExportTagged XML BibTex

New paper published in Computers and Operations Research

 The work On optimal participation in the electricity markets of wind power plants with battery energy storage systems  has been published in the journal Computers and Operations Research.  We present in this study a multi-stage stochastic programming model to find the optimal operation of a VPP in the day-ahead, intraday and secondary reserve markets hile taking into account uncertainty in wind power generation and clearing prices (day-ahead, secondary reserve, intraday markets and system imbalances). A case study with real data from the Iberian electricity market is presented. Preprint available at http://hdl.handle.net/2117/118479

Syndicate content