Implementation of the optimization problem for an energy management system
Visualitza/Obre
Estadístiques de LA Referencia / Recolecta
Inclou dades d'ús des de 2022
Cita com:
hdl:2117/110065
Tipus de documentProjecte Final de Màster Oficial
Data2017-06-15
Condicions d'accésAccés obert
Llevat que s'hi indiqui el contrari, els
continguts d'aquesta obra estan subjectes a la llicència de Creative Commons
:
Reconeixement-NoComercial-SenseObraDerivada 3.0 Espanya
Abstract
This master thesis project contributes to the development of an energy management system (EMS), whereby the thesis is part of a bigger project carried out by the Barcelona-based software company GreenPowerMonitor. The developed EMS coordinates the different components of a hybrid power plant by repeatedly solving an optimization problem over one day and computing the optimal setpoints of the components’ controllers for the next minutes. In the context of the EMS project, optimality is defined as maximizing the amount of used solar power while minimizing the battery losses for several future generation and consumption scenarios over the course of one day.
This thesis project focuses on examining ways to improve the computational costs of solving the optimization problem that lies at the core of the EMS and to implement the optimization problem in C++ as a “first draft” on its way towards a final product. Thereby, an already existing EMS prototype on the basis of Matlab/GAMS embedded in a simulation environment modeling a hybrid power plant in Cobija (Bolivia) is utilized. This prototype is translated into the open source modelling environment Julia/JuMP. The solver used throughout most of the project is the commercial mixed-integer linear problem (MILP) solver CPLEX. However, it is shown that the open-source solver CBC can be a serious alternative, especially for smaller power plants than the example case Cobija.
Test are conducted with the aim to qualitatively evaluate ways to keep the computation time of the solver within boundaries, while at the same time maintaining the quality of the results. As a measure of the EMS performance, the accumulated objective values over one day are compared. The conducted tests show, that setting a time limit to the solver, or extending the time to solve the problem, are good measures to keep the computation time within boundaries. The negative effects on the results are less pronounced than expected.
Furthermore, the optimization problem is implemented in the CPLEX solver specific application programming interface (API) in C++. The new C++ prototype is able to reproduce the expected results. In a future project, the input and output data processing will need to be implemented and refined.
Finally, an accumulated cost analysis is conducted that is able to show the economic utility of the EMS for a power plant operator.
TitulacióMÀSTER UNIVERSITARI EN ENGINYERIA DE L'ENERGIA (Pla 2013)
Col·leccions
Fitxers | Descripció | Mida | Format | Visualitza |
---|---|---|---|---|
Master_Thesis_Manuel_Ostertag.pdf | 1,771Mb | Visualitza/Obre |