Robust economic model predictive control of smart grids
(English) This thesis proposes a Robust Economic Model Predictive Control (REMPC) design based on deterministic approach for optimizing economic costs of energy production and dispatch in smart electrical grids. Robust Economic Model Predictive Control strategies for linear or linearized systems inc...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Fecha de publicación: | 2021 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/414771 |
| Acceso en línea: | https://hdl.handle.net/2117/414771 https://dx.doi.org/10.5821/dissertation-2117-414771 |
| Access Level: | acceso abierto |
| Palabra clave: | Àrees temàtiques de la UPC::Enginyeria elèctrica Àrees temàtiques de la UPC::Informàtica |
| Sumario: | (English) This thesis proposes a Robust Economic Model Predictive Control (REMPC) design based on deterministic approach for optimizing economic costs of energy production and dispatch in smart electrical grids. Robust Economic Model Predictive Control strategies for linear or linearized systems including algebraic constraints are developed. The contributions of this thesis are manifold: • The development of Robust Economic Model Predictive Control strategies for linear or linearized systems including algebraic constraints. • The development of Robust Optimization based EMPC strategies for the management of smart electrical grids incorporating several types of uncertainties. • The use of Robust Optimization technique to improve the robustness and reliability of the standard EMPC. • Assessing the efficiency and applicability of standard EMPC as well as its hierarchical and single-layer variants to solve economic energy dispatch problems of smart electrical grids. • The development and application of a novel constraints tightening minimax Robust EMPC method based on decomposing control inputs and states into dependent and independent components for tackling uncertain energy dispatch problems in smart electrical grids. • The development and application of a novel Robust EMPC Method based on Interval Arithmetic and Zonotopes from the perspective of Tube-based approaches for tackling uncertain energy dispatch problems in smart electrical grids. • The development of a Robust Economic Model Predictive Control strategy for linear system with algebraic constraints. The developed Robust EMPC strategies are categorized into three groups: single-layer EMPC, single-layer EMPC with pseudo-reference tracking, and hierarchical EMPC consisting of EMPC upper-layer and tracking MPC lower-layer. A generic control-oriented mathematical model of smart grids based on the theory of generalized flow-based networks is systematically developed. Additionally, a generic objective function for Economic MPC to solve economic energy dispatch of smart grids is also proposed. The developed mathematical model is repeatedly used to develop several Economic MPC formulations encompassing multi-objective cost functions for tackling uncertainties such as additive unexpected disturbances and energy costs fluctuation with periodically time-varying dynamics. First, the developed Economic MPC strategies are applied to smart micro-grid systems facing nominal periodic load demands and fixed energy prices, in order to assess recursive feasibility and stability of the system. After that, based on robust optimization techniques namely worst-case approach (minimax), the developed Economic MPC formulations are extended to tackle uncertainty of load demands, energy prices and intermittent energy sources. A closed-loop minimax control approach integrating the technique of affine dependence into the minmax approach for eliminating non-linearity and non-convexity of uncertain energy dispatch problems is also developed. Having discovered the computational complexity of minimax based on affine dependence technique, two novel constraints tightening Robust EMPC methods based on decomposition of control inputs and states into components for tackling uncertain energy dispatch problems in smart electrical grids are proposed. |
|---|