Defesa de Dissertação de Mestrado – José Diogo Forte de Oliveira Luna – 9/12/2019
06/12/2019 14:37
| Defesa de Dissertação de Mestrado | |
| Aluno | José Diogo Forte de Oliveira Luna |
| Orientador | Prof. Julio Elias Normey-Rico, Dr. – DAS/UFSC |
| Coorientador | Prof. Paulo Renato da Costa Mendes, Dr. – PPGEAS/UFSC |
| Data | 9/12/2019 (segunda-feira) – 8h00
Sala PPGEAS II (piso inferior) |
| Banca | Prof. Julio Elias Normey-Rico, Dr. – DAS/UFSC (presidente);
Prof. Marcelo Lobo Heldwein, Dr. – EEL/UFSC; Prof. Eduardo Camponogara, Dr. – DAS/UFSC; Prof. Marcelo de Lellis Costa de Oliveira, Dr. – DAS/UFSC. |
| Título | Contributions to MPC-Based Microgrid Central Controllers |
| Abstract: This work presents a set of contributions to microgrid central controllers (MGCCs) based on model predictive control (MPC), mainly focusing on providing them with voltage unbalance compensation and demand management (DM) capabilities. In a hybrid microgrid context, the presence of converters interfacing AC and DC buses can be used to tackle voltage unbalances, as long as the converters can control the negative sequence voltage of the bus they are connected to. However, unless each bus has a converter connected to it, it is necessary to establish a way to share the voltage unbalance compensation effort between the multiple converters in the microgrid. To address the challenge, two approaches are presented on this work, the first one being a novel convex formulation based on the negative sequence equivalent circuit of the grid, while the second one integrates the voltage unbalance compensation within a convex approximation of the optimal power flow problem. Concerning the demand response, in recent years, the usage of quality of experience (QoE) techniques has been getting attention, due to its ability to take into account the annoyance caused by DM actions to the user. While most of the proposals found in the literature, so far, only employ rule-based or fuzzy solutions, the present work integrates QoE metrics within the optimization problem solved by an MPC-based energy management system (EMS) for a smart home. Firstly, a survey was conducted using a digital questionnaire to assess the willingness of the local population to allow the EMS to interfere in their energy consumption pattern. The survey was conducted in Florianópolis and achieved a confidence interval of 95% and an error margin of 4.63%. The responses were used to leverage QoE curves describing the level of annoyance caused by the EMS interfering in several domestic appliances. These curves were used, then, to propose a QoE-aware demand response scheme on an MPC-based EMS. All of the techniques proposed in this work were tested by trustworthy simulations and provided promising results. | |
