EPE3301

Power Electronics
Control of Power Converters
Dr Lee Sze Sing, CEng

SzeSing.Lee@newcastle.ac.uk

SIT-NU Electrical Power Engineering

Basic Control Principles in Power Electronics
• Power converters are autonomous systems.
• Their performance does not only depend on the hardware
design but also on the control strategy used.
• It is important to understand some basic principles related to
the control of power converters.
• The standard approach to control power converters is to use
a linear controller to define a desired closed‐loop dynamic
along with a modulator to finally handle the power converter
switches.
• However, there are some control strategies that do not follow
this standard approach.

SS Lee 2
Power Electronics Control Targets
• Limited or constant switching frequency: The importance of
this control target is reflected in two main operation aspects
of the power converter. On one hand, limiting the switching
frequency results in known converter’s power losses, which
facilitates the hardware design based on the converter
efficiency and thermal dissipation. On the other hand, it
provides fixed current and/or voltage spectra, which
facilitates the interface filtering design.
• Reduced ripple: This helps to reduce the size of passive
components (capacitors and inductors) in a converter.
• High dynamic response: Power converters are normally part
of larger systems (e.g., electric vehicles, renewable power
plants), where a fast change in the power flow is required.
Therefore, a high dynamic response is crucial to satisfy those
requirements.

SS Lee 3
Power Electronics Control Targets
• No phase and amplitude error in the steady state for AC‐systems:
Standard controllers are normally designed to track constant
references (regulation problem). However, in AC‐systems, it is
required to track sinusoidal references. If the current or voltage
under control does not have the desired amplitude or phase, it can
lead to undesired effects. For instance, in a rectifier with unity
power factor, reactive power may be injected and/or absorbed by
the converter.

• Reduced harmonic content: Ideally for AC or DC applications,

currents or voltages under control are required to present only at
desired fundamental frequency or DC component with small
harmonic content, respectively. If the harmonic level is too high,
some detrimental consequences can be triggered. For instance, in
electrical drives, harmonics can lead to extra losses in the electric
motors. In grid‐connected applications, unwanted resonances in
the electricity grid might be triggered.

SS Lee 4
 Power Electronics Control Classification

(2) Controllers
without modulator

(1) modulation-base controllers

(3) Controllers with

embedded modulator

SS Lee 5
 Power Electronics Control Classification
• The controller has to track an average value of the
converter output (current, voltage, torque, power,
etc.).
• It will provide an average control input u.
• Since this average input cannot be directly
implemented in the converter, this is synthesized
by a modulation stage.
• The controller normally provides the required
duty‐cycle, u = d, to control the average value of y.
• To design a controller, an average model of the
converter and load is normally used, neglecting (1) modulation-base controllers
the modulator effect.
• Advantage of using a modulator comes from the
fact that it fixes the converter switching frequency
and produces a well‐defined spectrum over the
controlled variables.

SS Lee 6
 Power Electronics Control Classification
• Despite the well‐known advantages of using
modulators, there are some control strategies for
power converters that directly consider the state
of the switches (or output voltage levels) as
input, u = Sw.
• Hence, a modulation stage is not needed.
• This allows the control designer to use
instantaneous models of the power converters
and loads to design the controller.
• In that case, faster dynamic responses are
generally achieved when compared to using (2) Controllers without modulator
average models.
• However, the output variables present a spread
spectrum.

SS Lee 7
 Power Electronics Control Classification
• These control strategies explicitly consider
the modulator in their formulation along with
an instantaneous model of the converter and
load.
• To do this, the instants when the switches
have to commutate, tSw , within a sampling
time, Ts, are considered as control input, u =
tSw.
• The advantage of this control class is that a
fast dynamic is achieved during transients
with a fixed switching frequency and well‐
defined spectrum. (3) Controllers with embedded modulator

SS Lee 8
Basic Concept of Modulation
• The modulator’s role in controlling power electronic converters is the
same throughout each and every DC/DC, DC/AC, and AC/DC power
converters.

• Basically, the modulator decodes the controller’s effort in the form of

switching signals for driving the power semiconductors.

• PWM implementation can be done either in continuous‐time by using

analog signals or in discrete‐time as part of an analog‐to‐digital process

SS Lee 9
Modulation Methods for Voltage Source Inverters/Converters

SS Lee 10
Conventional Modulation Methods for Two‐Level
Voltage Source Inverters/Converters

SS Lee 11
Conventional Modulation Methods for Two‐Level
Voltage Source Inverters/Converters

SS Lee 12
 Space Vector Modulation (SVM)
Clarke’s Transformation (abc to αβγ transformation)
This transform ation projects the three-phase quantities onto tw o stationary axes
c c

α a α a
1 2
b β b β

c c

4 3
α a α a

b β b β
1 2 3 4 1 Rotating vector
SS Lee 13
Space Vector Modulation (SVM)
Voltage Vectors of a 2‐level Voltage Source Inverter

SS Lee 14
Space Vector Modulation
Synthesizing reference voltage vector
using inverter voltage vectors

SS Lee 15
 Selective Harmonic Elimination (SHE)
• Used in high‐power low‐switching frequency
applications.
• Use a precalculated switching pattern which
has been derived through the solution of a
system of nonlinear trigonometrical equations
in order to meet certain harmonic
requirements.

SS Lee 16
 Selective Harmonic Elimination (SHE)
• Commutation angles are predefined and precalculated in order to eliminate
low‐order harmonics and keep fundamental component tracking.
• To achieve this, the Fourier series of the predefined waveform is used to
equal each non‐desired low‐order harmonic to zero and in additionally
match the fundamental component with the desired modulation index given
by the reference.
• To explain the operating principle, consider the three‐angle case. The
Fourier series for the switched voltage waveform is given by

SS Lee 17
 Selective Harmonic Elimination (SHE)
• There are no even order harmonics due to half‐wave symmetry.

• Notice that the peak of ac components bn is a function of commutation angles.

• The fundamental component is forced to obtain the desired modulation index
Vˆ  MV ,V  V / 2 :
o ,1 max max dc 

SS Lee 18
 Selective Harmonic Elimination (SHE)
• There are no even order harmonics due to half‐wave symmetry.

• Notice that the peak of ac components bn is a function of commutation angles.

• Since there are no even order harmonics, and third‐order harmonics and its
multiples (called zero sequence harmonics, n = 3, 9, 15, etc) are cancelled by the
three‐phase connection of a balanced load, it is usual to eliminate the 5th and 7th
harmonic. This is achieved by replacing n = 5 and n = 7 in bn, and forcing the
coefficients to zero:

SS Lee 19
 Selective Harmonic Elimination (SHE)
• There are no even order harmonics due to half‐wave symmetry.

• The three angles and three equations form the nonlinear system to be solved.

SS Lee 20
 Selective Harmonic Elimination (SHE)
• To eliminate more harmonics, more angles can be added, but increasing the
complexity of the system.
• The general rule is that with k angles, k − 1 harmonics can be eliminated while
keeping control of the fundamental component.
• If more angles are considered, a natural choice for elimination would be the 11th,
13th, 17th and so on, since no even harmonics. There is no need to eliminate zero
sequence harmonics (n = 3, 9, etc) as they are eliminated through the three‐phase
connection of the load and do not appear in the line–line and load voltages.
• The set of equations cannot be solved online or analytically, being this the main
disadvantage of SHE. Hence, all the switching patterns have to be pre‐calculated
offline and stored in lookup tables.
• Many types of algorithms are used to solve these equations, mainly based on
iterative numerical techniques, such as genetic algorithms.

SS Lee 21
Selective Harmonic Elimination (SHE)
• Example 5‐angle solution for the entire modulation index and control
implementation

SS Lee 22
 Control of Power Converters
PWM‐Based Linear Control
external measurement noise or
The main target is to design disturbance system uncertainties
a feedback controller, C(s),
to reduce the tracking
error, e = r – y, by applying
a suitable control input, u,
to the system Go(s).

The output of the linear

controller normally
provides a duty‐cycle, d,
which has to be synthesized
by a modulator.

Linear feedback controller: (A) general closed‐loop

representation and (B) closed‐loop including modulator.

SS Lee 23
 Control of Power Converters
PWM‐Based Linear Control
external measurement noise or
The actual input system uncertainties
disturbance
applied to the system, given
by the state of the power
switches u = Sw, differs
from the nominal input
provided by the controller,
d. This mismatch can be
interpreted as an input
disturbance given by ηi = Sw
– d (modulation noise).

In digital implementations,
the controller sampling is
normally synchronized with
the PWM carrier. Linear feedback controller: (A) general closed‐loop
representation and (B) closed‐loop including modulator.

SS Lee 24
Control of Power Converters
Example PWM‐Based Linear Control

Current control with SVM

SS Lee 25
Load current
 Control of Power Converters
Control without Modulator: Hysteresis Current Control
• The basic idea of hysteresis
current control is to keep the
current inside the hysteresis
band by changing the
switching state of the
converter each time the
current reaches the boundary.
  iL*  iL
  positive  increase iL
  negative  decrease iL
• How to increase or decrease
current?
diL
vL
dt
v  Vdc  iL increase
v  Vdc  iL decrease

SS Lee 26
 Control of Power Converters
Control without Modulator: Hysteresis Current Control (3‐phase)
• The performance of the
hysteresis controller is good,
with a fast dynamic response.
• The switching frequency
changes according to variations
in the hysteresis width, load
parameters, and operating
conditions. This is one of the
major drawbacks of hysteresis
control, since variable switching
frequency can cause resonance
problems.
• When implemented in a digital
control platform, a very high
sampling frequency is required
in order to keep the controlled
variables within the hysteresis
band all the time.

SS Lee 27
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
FCS‐MPC uses a model of the system to predict the future behavior of the
variables until a predefined horizon in time, and selection of the optimal
actuations by minimizing a cost function. This structure has several
important advantages:
• Concepts are very intuitive and easy to understand.
• It can be applied to a great variety of systems.
• The multivariable case can be easily considered.
• Dead times can be compensated.
• Easy inclusion of non linearities in the model.
• Simple treatment of constraints.
• The resulting controller is easy to implement.

SS Lee 28
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
In the design stage of FCS‐MPC for the control of a power converter, the
following steps are identified:
• Modeling of the power converter identifying all possible switching states
and its relation to the input or output voltages or currents.
• Defining a cost function that represents the desired behavior of the
system.
• Obtaining discrete‐time models that allow one to predict the future
behavior of the variables to be controlled.

SS Lee 29
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Converter modelling
• For modelling of converters, the switching states or voltage levels are
considered for single‐phase system. For three‐phase systems, the voltage
vectors are considered.
• For example, in a three‐phase, two‐level converter, the eight switching
states generate seven different voltage vectors, with two switching states
generating the zero vector.

SS Lee 30
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Cost function
• Each different application imposes several control requirements on the
systems such as current control, torque control, power control, low
switching frequency, etc. These requirements can be expressed as a cost
function to be minimized.
• The most basic cost function to be defined is some measure of error
between a reference and a predicted variable, for example, load current
error, power error, torque error, and others.
• One of the advantages of the predictive control methods is the possibility
to control different types of variables and include restrictions on the cost
function.
• In order to deal with the different units and magnitudes of the controlled
variables, each term in the cost function is multiplied by a weighting
factor that can be used to adjust the importance of each term.

SS Lee 31
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
System discrete‐time model for prediction
• When building the model for prediction, the controlled variables must be
considered in order to get discrete‐time models that can be used for the
prediction of these variables. For example, for controlling load current, a
discrete‐model describing the relationship of load current and converter
voltage vectors is required.
• To get a discrete‐time model it is necessary to use some discretization
methods. For first‐order systems, it is simple to approximate the
derivatives using the Euler forward method, that is, using

where Ts is the sampling time.

SS Lee 32
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Implementation
When implemented, the controller must consider the following tasks:
• Predict the behavior of the controlled variables for all possible switching
states (or voltage vectors).
• Evaluate the cost function for each prediction.
• Select the switching state that minimizes the cost function.

SS Lee 33
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter

Predictive current control algorithm

SS Lee 34
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter

This control strategy can be summarized in the following steps:

• Build a model of the converter and its possible switching states.
• Define a cost function g.
• Build a discrete‐time model of the load for prediction.

SS Lee 35
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a model of the converter and its possible switching states.

Clarke’s Transformation:

SS Lee 36
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a model of the converter and its possible switching states.

Switching states and voltage vectors Voltage vectors in complex plane

SS Lee 37
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Define a cost function g.
The objective of the current control scheme is to minimize the error between the
measured currents and the reference values.
The cost function is expressed in orthogonal coordinates and measures the error
between the references and the predicted currents.

where and are the real and imaginary parts of the predicted
load current vector, for a given voltage vector. The reference
Currents and are the real and imaginary parts of the reference
current Vector . For simplicity, we will assume that this reference
current does not change sufficiently in one sampling interval, so we will consider

SS Lee 38
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a model of the load for prediction.

The equations for load current dynamics for each phase can be written as

SS Lee 39
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a model of the load for prediction.
Vector equation for the load current dynamics can be obtained by applying
Clarke’s transformation:

SS Lee 40
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a model of the load for prediction.
Considering the space vector definition for the inverter voltage given by
Clarke’s transformation, the following definitions for load current and back‐
emf space vectors can be written:

Then the load current dynamics can be described by the vector differential
equation

where v is the voltage vector generated by the inverter, i is the load current
vector, and e the load back‐emf vector.
SS Lee 41
 Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter
• Build a discrete‐time model of the load for prediction. (Discretization)
The discrete‐time model will be used to predict the future value of
load current from voltages and measured currents at the kth sampling instant.
The load current derivative di/dt is replaced by a forward Euler approximation.

The back-emf can be calculated by considering measurements of the load voltage and
current with the following expression:
SS Lee 42
 Control of Power Converters
Control without Modulator:
Finite Control Set Model
Predictive Control (FCS‐MPC)
Example: Predictive current control of a
two‐level three‐phase inverter
• Control algorithm

Example values of
cost function for
each voltage
vector in one
sampling period.

SS Lee 43
Control algorithm
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter

SS Lee 44
Control of Power Converters
Control without Modulator: Finite Control Set Model Predictive
Control (FCS‐MPC)
Example: Predictive current control of a two‐level three‐phase inverter

Load current in
time domain

Sampling period = 100 us

Performance can be
improved by increasing
sampling frequency.
However, the switching
frequency is also
increased that results in
higher switching loss.
Sampling period = 25 us
SS Lee 45