Decentralized optimization for energy efficiency under stochasticity

Abstract : New energy systems are designed to absorb a large share of renewableenergy in a decentralized fashion. Their optimized management raises specificissues. We study mathematical formulation as large scale multistagestochastic optimization problems. We focus on time and space decompositionmethods in a stochastic setting.In the first part of this manuscript, Time decomposition inoptimization and management of home microgrids, we apply stochasticoptimization algorithms to the management of small scale microgrids. We compare different optimization algorithms on two examples:a domestic microgrid equipped with a microCombined Heat and Power generator and a battery;a domestic microgrid equipped with a battery and solar panels.In the second part, Mixing time and spatial decomposition inlarge-scale optimization problems, we extend the previous studies tolarger microgrids, where different units and storage devices are connected together. As a direct resolution by Dynamic Programming of such large scale systemsis untractable, we propose original algorithms mixing time decomposition on the one hand, and price and resource spatial decomposition on the other hand.In the third part, Contributions to Stochastic Dual Dynamic Programming,we focus on the Stochastic Dual Dynamic Programming (SDDP) algorithm,a well-known algorithm to solve multistage stochastic optimizationproblems. We present a new stopping criteria based on a dual versionof SDDP which gives a deterministic upper-bound for the primal problem
Document type :
Theses
Complete list of metadatas

Cited literature [104 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/tel-02134163
Contributor : Abes Star <>
Submitted on : Monday, May 20, 2019 - 12:56:54 PM
Last modification on : Thursday, July 11, 2019 - 10:50:05 AM

File

TH2018PESC1147.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02134163, version 1

Collections

Citation

François Pacaud. Decentralized optimization for energy efficiency under stochasticity. Optimization and Control [math.OC]. Université Paris-Est, 2018. English. ⟨NNT : 2018PESC1147⟩. ⟨tel-02134163⟩

Share

Metrics

Record views

107

Files downloads

132