Model Predictive Control of microturbine-based distributed generation systems in the intelligent operation of distribution networks
This repository contains all MATLAB simulations (Simulink) of my final master thesis at the University of Zanjan. Since micro-turbine generation (MTG) systems generate high-frequency voltages and currents, precise control of interface converters plays a crucial role. In this project, I tested various interface converters (i.e., AC-DC-AC and AC-AC) with/without battery storage, and two main types of control methods were tested, including traditional ones and Model Predictive Control (MPC). The results showed that MTG with MPC has higher flexibility, and its performance improved significantly. Moreover, for providing secure power on the demand side, energy storage is considered to save excessive amounts of power. [For reading my papers, there is a link at the end of this page.].
The main novelties of this project can be summarized in these points:
- a comprehensive analysis of MTG with different interface converters and the role of storage in such a system.
- applying MPC to MTG's system for controlling interface converters as a new approach.
- applying MPC to MTG for controlling fuel and speed as a novel method.
- using a new filter configuration (called Remover Ripple Circuit (RRC)) on the MTG's output for obtaining sinusoidal voltage and current.
Microturbine (MT) generation (MTG) system is a relatively new Distributed Generation (DG) and fast-growing technology that is appropriate for small-scale generation because of its compact size, quick start, long lifetime, reliability and durability, low initial and maintenance costs, low emission level, and ability to operate with alternative fuels, including natural gas, biogas, and diesel. The MTG is indeed a small gas turbine that operates based on the Brayton thermodynamic cycle. Smaller size and higher power density are desirable reasons to use MT in low-power demand areas instead of conventional gas turbines. The MT generates power within the range of 25–1000 kW, and its efficiency is around 20–30%, where up to 80% can also be achieved by Combined Heat and Power (CHP) system. These generation systems are used for a wide range of applications, such as peak shaving, CHP, remote power, premium power, transportation systems, resource recovery, baseload power, and standby services. For more information, read our chapter: https://www.sciencedirect.com/science/article/pii/B9780128042083000042
All the simulations are carried out in MATLAB software. In the repository, you can find three different folders, each consisting of various Simulink files.The AC-DC-AC folder means the MTG is connected to the load or grid through a double conversion interface. In such interfaces, at first, high-frequency AC voltage is converted to DC voltage, and then, through an inverter, it is converted to AC voltage with a 50/60 Hz frequency and suitable amplitude. In this folder, the rectifier may be uncontrolled (diode rectifier folder) or controlled (thyristor rectifier folder). Various scenarios are tested for each type, including off-grid, on-grid, both modes, and with or without storage.In AC-AC mode, the simulations are carried out with a cyclo converter. The results of AC-AC are not as good as double conversion due to the high-frequency voltage from the MTG side. There are some papers about working on a matrix converter, and you can read and clone them.The MPC folder consists of files in which this method is used for controlling MTG's inverter in various situations. Last but not least, the file with storage, RRC, or filter postfixes means simulation with storage and a new RRC filter, respectively.
- All simulation files were already created (2014–2016), and after that, I just published papers. Unfortunately, I had no time to update them. Before posting this repository, I tested all of them, and the files worked without error. If you encounter any errors, especially in MPC files, let me know in the issue section.
- For more information about this project and papers, please visit the project page in ResearchGate