Model predictive control of grid-connected voltage source converters

Yoo, D 2013, Model predictive control of grid-connected voltage source converters, Doctor of Philosophy (PhD), Electrical and Computer Engineering, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Yoo.pdf Thesis application/pdf 18.40MB
Title Model predictive control of grid-connected voltage source converters
Author(s) Yoo, D
Year 2013
Abstract In this thesis, the main focus is on the design and implementation of an advanced control scheme, namely model predictive control (MPC) to the grid- connected voltage source converter (VSC) for a three phase system. MPC is a control paradigm that solves a mathematical optimization problem based on a dynamic model of the system. Due to the computationally demanding nature of MPC, the areas of applications have long been restricted to slow dynamical systems. However, with the recent advancement of microprocessor and simu- lation technologies, application of MPC is now even possible for the control of power electronics. With a very powerful concept such as on-line cost optimisation, input/output constraint handling and model-based design, MPC is able to offer the optimal actuation that allows one to achieve very fast dynamics, while also considering uncertainties such as system parameter variations and unknown disturbances. Furthermore, it is also possible to take advantage of the discrete nature of the power converters and choose from the possible switching states the optimal solution according to the minimization of a predefined cost.  Exploring these advantages of MPC and making them suitable for the control of power converters are the key focus of the thesis.

The first part of the thesis investigates a multi-variable control scheme, namely a predictive voltage controller that controls both DC bus voltage and re- active current (i.e. q-axis current) in the synchronous reference frame. Explicit tuning methods of MPC are introduced to improve the closed-loop transient response as well as improving the robustness against the parameter variations such as the grid inductance.

The second part of the thesis focuses on the predictive current control design. A predictive current controller for VSC with LCL (inductor-capacitor- inductor) input filter is first proposed with a robust control scheme that employs nominal and disturbance rejection control parts. The nominal control part is designed using the reduced-order model (i.e. L filter model) to control dominant dynamics of the LCL filter where as the disturbance rejection control part actively suppresses the disturbance due to unmodeled dynamics of LCL filter (i.e. resonance of the LCL filter). Following from this, a predictive resonant controller is presented to control the converter in the stationary frame axis. A resonant module with a grid frequency is embedded in the model to handle the periodicity in the measured states and the reference inputs. The proposed de- sign considers the periodic input constraints in the stationary frame as well as disturbances due to grid voltage distortion. The last part of the thesis investigates the stability aspect of a finite control set predictive control (FCS-MPC) method and presents a design framework to handle the imposed the output current constraints in the cost function.

All of the presented control methods in this thesis are experimentally validated on a 1kW prototype converter that has been built by the author.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Electrical and Computer Engineering
Keyword(s) MPC
PWM converter
Robust
Versions
Version Filter Type
Access Statistics: 665 Abstract Views, 722 File Downloads  -  Detailed Statistics
Created: Fri, 02 May 2014, 15:55:58 EST by Lynne Johns
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us