DAB_1714901843

FACULDADE DE E NGENHARIA DA U NIVERSIDADE DO P ORTO
Dual Active Bridge Converter for

Electric Vehicle Charging
André Filipe Gama de Sousa Rodrigues
Mestrado em Engenharia Eletrotécnica e de Computadores
Supervisor: António José de Pina Martins
October 29, 2023

© André Filipe Gama de Sousa Rodrigues, 2023
Abstract
Electric vehicles (EVs) are gaining popularity as a sustainable transportation option, but their
widespread adoption is limited, due to the availability and efficiency of charging infrastructures,
among other factors. There are different charging solutions, in particular with regard to electronic
converters connected to the battery.
The dual active bridge (DAB) converter is a promising technology for EV charging, as it
can operate at high power levels, offer high efficiency and due to its bidirectional characteristics.
Therefore its understanding and evolution can be a key factor in expanding the market share of
electric vehicles.
In this dissertation, a comprehensive study of the DAB converter for electric vehicle (EV)
charging applications is presented, including its behavior, main characteristics, and type of con-
trol. A DAB prototype for EV charging will be designed and implemented to analyse its per-
formance through simulation, and be further evaluated through experimental testing. The goal
is to demonstrate the potential of DAB converters to enable efficient, powerful and reliable EV
charging.
i
ii
Resumo
Os veículos elétricos (EVs) estão a ganhar popularidade como uma opção de transporte susten-
tável, mas a sua adoção generalizada é limitada, sendo a disponibilidade e eficiência das infraestru-
turas de carregamento um dos fatores. Existem diversas soluções de carregamento, em particular
no que respeita ao conversor eletrónico com ligação à bateria.
O conversor "dual active bridge" (DAB) é uma tecnologia promissora para o carregamento
de veículos elétricos, pois pode operar em altos níveis de potência, oferecer grande eficiência e
devido às suas caracteristicas bidirecionais. Portanto, o seu entendimento e evolução podem ser
um fator chave na expansão da quota de mercado dos veículos elétricos.
Nesta dissertação, é apresentado um estudo abrangente do conversor DAB para aplicações de
carregamento de veículos elétricos (EV), incluindo o seu comportamento, principais caracterís-
ticas e tipo de controlo. Também é projetado e implementado um conversor DAB para carrega-
mento de veículos elétricos e é analisado o seu desempenho por meio de simulação. Além disso,
é analizado o seu desempenho em protótipo por meio de testes experimentais verificando o seu
comportamento com hardware. O objetivo será demonstrar o potencial dos conversores DAB para
permitir um carregamento de veículos elétricos eficiente, potente e confiável.
iii
iv
Contents
1 Introduction 1
1.1 Context and Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 State of the Art 3

2.1 World Electricity Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Electric Vehicle Shift and EV Charging Stations . . . . . . . . . . . . . . . . . . 3
2.2.1 AC vs DC charging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2 Levels of EV charging . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Vehicle-to-Grid Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Topology of Bidirectional DC-DC Converters . . . . . . . . . . . . . . . . . . . 8
2.4.1 Isolated Converters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4.2 Control Schemes for Bidirectional Converters . . . . . . . . . . . . . . . 12
3 Background on the Dual Active Bridge 17

3.1 Theoretical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4 Interleaved Dual Active Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 Simulation Test 23
4.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Specifications and Circuit Setup . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.3.1 Phase Shift Control Implementation . . . . . . . . . . . . . . . . . . . . 26
4.3.2 PI controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.4.1 30º vs 72º Phase Shift at Permanent Regime . . . . . . . . . . . . . . . . 28
4.4.2 Reverse Flow in Permanent Regime . . . . . . . . . . . . . . . . . . . . 30
4.4.3 PI Control Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.4.4 Step responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.4.5 Interleaved DAB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
v
vi CONTENTS
5 Experimental Test 41
5.1 Microcontroller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.1.1 Square Wave Implementation . . . . . . . . . . . . . . . . . . . . . . . 41
5.2 Transistor Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.3 Circuit Build and Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.4.1 Preliminary Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.4.2 DAB testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6 Conclusions and Future Work 59

6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
References 61
List of Figures
2.1 World electricity generation share in 2020 by fuel category and by fuel [1]. Data:
BP Statistical Review of World Energy 2021 . . . . . . . . . . . . . . . . . . . . 4
2.2 Global greenhouse gas emissions by sector [2] . . . . . . . . . . . . . . . . . . . 5
2.3 AC vs DC charging diagram. [3] . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.4 DC fast charge charging curve. Power output as y-axis and State of Charge (SoC)
as x-axis. [3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.5 Comparison of Volkswagen ID.4 fast charging curve (150+ kW) with normal
charging curve (50 kW) [4] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.6 Diagram of bidirectional EV charging stations in grid [5] . . . . . . . . . . . . . 9
2.7 DC-DC topology flowchart [5] . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.8 Bidirectional Flyback DC-DC Converter [5] . . . . . . . . . . . . . . . . . . . . 10
2.9 Isolated Cuk & Sepic/Zeta Bidirectional DC-DC Converter [5] . . . . . . . . . . 10
2.10 Push-Pull Bidirectional DC-DC Converter [5] . . . . . . . . . . . . . . . . . . . 11
2.11 Forward Bidirectional DC-DC Converter [5] . . . . . . . . . . . . . . . . . . . . 11
2.12 Dual Half-Bridge Bidirectional DC-DC Converter [5] . . . . . . . . . . . . . . . 12
2.13 Half Bridge-Full Bridge Bidirectional DC-DC Converter [5] . . . . . . . . . . . 12
2.14 PID control feedback diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.15 Sliding mode control feedback diagram . . . . . . . . . . . . . . . . . . . . . . 13
2.16 Fuzzy control feedback diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.17 Digital control feedback diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1 DAB converter circuit [6] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.2 DAB waveforms for forward power flow: Primary Voltage, Secondary Voltage and
Coil Current respectively [6] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 DAB Current flow at each operating mode. . . . . . . . . . . . . . . . . . . . . . 19
3.4 PI controller diagram for DAB applications . . . . . . . . . . . . . . . . . . . . 20
3.5 Waveforms for (A) EPS and (B) DPS [7] . . . . . . . . . . . . . . . . . . . . . . 22
4.1 DAB circuit on PSIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.2 Interleaved DAB circuit on PSIM. . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3 Phase shift control Circuit on PSIM. . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4 Phase shifted PWM wave built step by step. . . . . . . . . . . . . . . . . . . . . 27
4.5 Phase shift control circuit for interleaved DAB on PSIM. . . . . . . . . . . . . . 27
4.6 30 degree phase shifted waves for first and second DAB. . . . . . . . . . . . . . 28
4.7 PI Controller Circuit on PSIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.8 C block code for power charging curve. . . . . . . . . . . . . . . . . . . . . . . 29
4.9 PSIM power charging curve graph. . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.10 Waveforms for 30 degree phase shift. . . . . . . . . . . . . . . . . . . . . . . . . 32
vii
viii LIST OF FIGURES
4.10 Waveforms for 30 degree phase shift. . . . . . . . . . . . . . . . . . . . . . . . . 33

4.11 Waveforms for 72 degree phase shift . . . . . . . . . . . . . . . . . . . . . . . . 34
4.11 Waveforms for 72 degree phase shift . . . . . . . . . . . . . . . . . . . . . . . . 35
4.12 Waveforms for a negative 30 degree phase shift. . . . . . . . . . . . . . . . . . . 36
4.12 Waveforms for a negative 30 degree phase shift. . . . . . . . . . . . . . . . . . . 37
4.13 Waveforms for the charging power and angle phi PI response. . . . . . . . . . . . 37
4.14 Overview of output current against reference current. . . . . . . . . . . . . . . . 38
4.15 Zoomed waveforms for the output current step response. . . . . . . . . . . . . . 39
4.16 Waveforms for interleaved DAB with 30 degree shift. . . . . . . . . . . . . . . . 40
5.1 STM32F429 Development Board. . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.2 STMCubeIDE timer interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.3 Centre align mode principle, [8]. . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.4 Centre align mode comparison to edge aligned, [9]. . . . . . . . . . . . . . . . . 44
5.5 Output Compare with active level on match. . . . . . . . . . . . . . . . . . . . . 45
5.6 HAL library code for signal declaration . . . . . . . . . . . . . . . . . . . . . . 46
5.7 CD4050B CMOS Hex Inverting Buffer . . . . . . . . . . . . . . . . . . . . . . 46
5.8 Single phase inverter with current sensors. . . . . . . . . . . . . . . . . . . . . . 47
5.9 Coil and transformer series circuit. . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.10 Connector pins and bridge circuit diagram. . . . . . . . . . . . . . . . . . . . . . 48
5.11 Full DAB circuit build. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.12 PWM signal confirmation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.12 PWM signal confirmation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.13 Full bridge and diode bridge test result. Yellow: PWM; magenta and light blue:
primary and secondary voltages; dark blue: primary AC current. . . . . . . . . . 52
5.14 Testing results for DAB at 30º angle phase shift and load voltage of 24 V. . . . . 54
5.17 Testing results for DAB at 0º angle phase shift and load voltage of 24 V. . . . . . 57
List of Tables
2.1 Differences between AC and DC charging. . . . . . . . . . . . . . . . . . . . . . 7

2.2 Summary of control schemes in bidirectional converter [5] . . . . . . . . . . . . 15
3.1 Operating modes of the DAB . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.1 Simulation value table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1 Internal Trigger Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
ix
x LIST OF TABLES
Abbreviations and Symbols
AC Alternating Current
DAB Dual Active Bridge
DC Direct Current
DPS Double Phase Shift
EPS Extended Phase Shift
ESS Energy Storage System
EV Electric Vehicle
G2V Grid to Vehicle
IGBT Insulated Gate Bipolar Transistor
MOSFET Metal Oxide Semiconductor Field Effect Transistor
PHEV Plug-in Hybrid Electric Vehicle
PWM Pulse-Width Modulation
SoC State of Charge
SPS Single Phase Shift
TPS Triple Phase Shift
V2G Vehicle to Grid
ZVS Zero Voltage Switching
iL Current in coil
Io Current output
Po Current input
Vo Voltage output
Vp Voltage in primary winding
Vs Voltage in secondary winding
Vi Voltage input
φ Phase shift
xi
Chapter 1
Introduction
1.1 Context and Motivation
Electric vehicles (EV) have emerged as a promising solution to mitigate the environmental impacts
of transportation, reduce dependence on fossil fuels, and promote sustainable mobility. In the
past few years, governments, manufacturers and consumers are increasingly embracing EVs as a
viable alternative to internal combustion engine vehicles. However, despite the numerous benefits
of electric vehicles, there are still some challenges and limitations that hinder their widespread
adoption. Their limited amount of energy available, which is strictly dependent on the capacity
of the vehicle’s battery, affects the range of the EV and setbacks its use for long distance travel.
Another hindrance is the limited availability of charging stations for public use, as the transition
to widespread EV adoption is not solely dependent on the advancements in vehicle technology but
also relies heavily on the development of robust charging infrastructure and charging technology.
The adoption of electric vehicles continues to grow and it is important to ensure that charging
infrastructure is able to support the increasing demand for electricity and do so in a way that is
efficient, reliable, and sustainable. To address these challenges, researchers and engineers are
continuously exploring advanced charging technologies and power electronics solutions.
The Dual Active Bridge (DAB) converter is one such technology that has gained prominence
in the field of EV charging due to it’s bidirectional capability, among with its high power ratings
and energy density, galvanic isolation and efficiency, and it is most commonly used in level 2 and
level 3 charging stations.
1.2 Objectives
This dissertation takes a closer look at these charging stations, focusing on understanding the the-
ory and functionalities of the DAB in the context of electric mobility. The theory is subsequently
put into practice with software test simulations, emulating its role in a real DC fast charger, and
1
2 Introduction
analysis of the waveforms and properties of some relevant variables produced at different con-
ditions. Further tests are executed using hardware components for a "hands on" experience to
analyse the working of the converter in real life situations in a scaled environment.
1.3 Structure
In addition to the introduction, this document is organised into five more chapters:
• Chapter 2 includes a review of the literature on the topic and an overview of the current state
of the art in regards to electric vehicle charging and the converters that rival the DAB.
• Chapter 3 provides the necessary background information on the functionalities and main
characteristics of the DAB converter.
• Chapter 4 details the PSIM simulation’s implementation and results.
• Chapter 5 shows the experimental tests made, their results, the hardware that was used and
explains the circuits built.
• Chapter 6 presents the conclusions drawn from the work completed and suggest some future
works related to this one.
Chapter 2
State of the Art
2.1 World Electricity Production
When analysing data and studies from the global perspective of electricity production in the BP
statistical review of world energy 2021, Figure 2.1, updated in January 2022, it appears that about
62% of global energy is generated by fossil fuels. Of the remaining percentage, 10% is gen-
erated by nuclear plants, 1% via other forms of fuel and leaving 28% created from renewable
sources [10]. This shows that fossil fuels are still, nowadays, the main resource used in the pro-
duction of electric energy in the world. As a type of non-renewable resource and due to its high
demand, the continued reliance on fossil fuels is clearly unsustainable in the long term. This raises
concern over the sustainable development of the world. However, this is just one of the main rea-
sons that motivate the need to stop using these non-renewable resources. Other reasoning include
the increase of pollution in the environment which this type of energy production promotes. A ma-
jor problem of fossil fuels is the release of greenhouse gases during its combustion. These gases
boosts the constant climate change in the terrestrial ecosystem, progressively and irreversibly de-
stroying the planet.
To address the problems associated with fossil fuels, the world has increasingly focused on
energy production through renewable energy sources. These types of sources use various natural
phenomena to achieve a “greener” energy production, free from greenhouse gases, pollution and
their consequences to the environment. As they are constantly present in nature, these resources
are inexhaustible and thus promote the planets sustainable development. Various studies indicate
that the production of energy by only using renewable sources, together with its efficient storage,
can manage to suppress all the world’s energy needs, with affordable prices, wide availability and
with a clean and sustainable nature [11].
2.2 Electric Vehicle Shift and EV Charging Stations
The world emits each year approximately 50 billion tonnes of green house gases, and as seen
in figure 2.2, 12% are due to road transportation [2]. These vehicles are constantly releasing
3
4 State of the Art
Figure 2.1: World electricity generation share in 2020 by fuel category and by fuel [1]. Data: BP
Statistical Review of World Energy 2021
carbon dioxide (major component for green house gases) into the atmosphere and are responsible
for around 25% of the total carbon dioxide emissions in the European Union [12]. This number
explains the rise in rigidity of the world’s commissions, in reaching carbon neutrality, in the so-
called "decarbonisation" strategy by 2050 [13].
Due to all these factors, there is a huge push for the the replacement of combustion vehicles
with electric vehicles [14]. Electric Vehicles, Hybrid Electric Vehicles and Plug-in Hybrid Electric
Vehicles use electrical energy either as a primary fuel or to improve the efficiency of conventional
combustion engines. They all use a battery, usually lithium, to store the electrical energy needed
to power the engine. A very sudden spread of these vehicles is expected in the coming years, as
their cost (purchase, use, maintenance, etc.) will become similar or better compared to cars with
a combustion engine. This means that charging stations must develop at a rate at least equal to the
rate of sales of these vehicles. Due to this fact and with the evolution of the Smart Grid, the storage
of energy for later charging of these vehicles, through the Grid to Vehicle (G2V) technology, which
is usually made up of a cascaded connection of an AC-DC and a DC-DC converter, is one of the
most appreciated solutions. This technology allows the vehicle’s battery to be in parallel with
the electrical network during charging operations. However, it also allows the vehicle to supply
energy to the grid like any other storage system (V2G mode), using bidirectional converters (power
electronics) that guarantee an adequate strategy for managing the power flow.
2.2.1 AC vs DC charging
The power that is provided by the electric grid is always AC power, however, an electric vehicle, or
any other electric component by that matter is only capable of storing DC power. This factor urges
the power to be converted before storing it in the EV. There are two ways to charge the vehicle:
2.2 Electric Vehicle Shift and EV Charging Stations 5
Figure 2.2: Global greenhouse gas emissions by sector [2]
AC and DC charging, each having their benefits and disadvantages and with its main difference
being in the location of the conversion of power [3], in Figure 2.3 this difference can be observed.
In AC charging the power enters the car as AC, and it is only converted to DC in an onboard
charger by an AC-DC converter that each EV possesses. These onboard chargers limits the amount
of power converted, which makes the speed of charge of AC charging very time-consuming and
much inferior in comparison to DC charging. In DC charging, or DC ’fast’ charging as it is also
referred, the DC power is directly passed to the battery, with no need for the onboard charger.
This charger possesses converters with far greater efficiency and much more powerful, so that the
car is delivered a big amount of power for the quickest possible time of charge. Although big
power dump to the car may improve the charging time, its battery may be slightly damaged in
comparisons to using AC charging in the long run, in recent studies on AC and DC chargers it
was found that cars who used ’fast’ charging more than three times a month would experience
a 10% increase in battery degradation [3][15]. Due to this possible damage and to the different
power threshold different EVs can withstand, each car has its own charging curve for DC fast
charge. This curve will also affect the total charging time. The charging station is in constant
communication with the EV and once this connection commences, the charger quickly reaches its
maximum power output, however it is usual for an EV to only charge at it’s maximum rate for a
6 State of the Art
fraction of the full charging period. After this, the power provided to the vehicle slowly decreases
as the battery gets more charged, and when it reaches 80% a faster decrease of power starts. In
Figure 2.4 it is shown a model of a common DC fast charge charging curve, with power output in
the y-axis and SoC in the x-axis. A real charging curve is shown in Figure 2.5 which displays the
charging curve of a Volkswagen model ID.4, which can withstand 130 kW on fast charge. This
diagram shows the curve with a 50 kW charger in comparison to a 150 kW or greater charger.
Table 2.1 shows further differences between both charging types.
Figure 2.3: AC vs DC charging diagram. [3]
Figure 2.4: DC fast charge charging curve. Power output as y-axis and State of Charge (SoC) as
x-axis. [3]
2.2 Electric Vehicle Shift and EV Charging Stations 7
Figure 2.5: Comparison of Volkswagen ID.4 fast charging curve (150+ kW) with normal charging
curve (50 kW) [4]
AC charging DC charging
Position of
Onboard charger Inside charging station
power converter
Speed of charge Slower charging compared to DC Faster charging
Placement of Common for house-holds Only used for public
chargers and public charging charging
Charging curve Constant line Degrading charging curve
DC fast charge for a long
Electric battery Safe for car battery
time may degrade battery
Table 2.1: Differences between AC and DC charging.
2.2.2 Levels of EV charging
EV charging is also separated in three levels, these are used to differentiate the range of power
output that is used to charge the car [16] [17].
• Level 1: The slowest charging level in terms of time, with an estimated charging time of
between 22 hours up to 40 hours (estimation for an EV with a 40 kWh battery). It charges
the car with AC power ranging between 1 kW to 1.8 kW.
• Level 2: Similar to the first level, Level 2 also charges the vehicle with AC power, however
with a big increase in power output (3 kW - 22 kW), with the times of charging of a 40 kWh
battery from 2 hours up to 13 hours.
8 State of the Art
• Level 3: The last level is also known as DC fast charging, and as it’s name suggests it
provides DC power to the vehicle. This level of charging is used for public charging, and
the typical power output can go as high as 360 kW, which depending on the vehicle charge
acceptance range the time of charge can be cut to only 15 minutes for a 40 kWh battery.
2.3 Vehicle-to-Grid Technology
A concept first put forward in 1997 [18], V2G consists on the ability for EVs to supply energy
back to the electric grid, creating a bidirectional power flow between both [19].
EV batteries have the capability to store a substantial amount of power, and when parked (a car
is parked for about 95% of their life span) could be perceived as a sizeable untapped energy storage
device, and if used to its full potential could offer enormous benefits to the electric grid such as
reactive power support, active power regulation, reducing electricity transport losses [20][21] and
could benefit also our own household consumption. It is estimated that an average Portuguese
family household consumes 275 kWh per month [22], dividing this number by 30 days gives a
consumption per day of about 9 kWh, this means that an average EV car battery of 40 kWh could
power a family home for about 4 days, a number that can be increased as these car batteries may
even go to as high as 100 kWh.
Despite knowing this, using electric vehicles to support the electric grid is no simple task
as there are structures that need to be put in place in order to fulfill it’s bidirectional role. The
smart grid must be able to support such system as well as EV charging stations and the vehicles
itself must allow for the power to flow in either way. In regards to DC fast charging stations, the
implementation of bidirectional DC-DC converters is a must. Figure 2.6 shows the interaction
between DC-DC converters in EV charging stations and the smart grid. By interacting power
sources and energy storage devices with the bidirectional feature, a reduction in size and cost as
well as an increase in efficiency and performance of the system overall is reached as there is no
need for two different converters to control each direction of flow [5].
2.4 Topology of Bidirectional DC-DC Converters
Bidirectional DC-DC converters can be divided between isolated and non-isolated, these cate-
gories mainly diverge, as the name refers, in their isolation. Non-isolated converters have no
magnetic isolation when transferring the power as they do not use transformers. This topology is
simpler to arrange, lighter and do not experience magnetic interference. Isolated converters on the
other hand, convert the DC voltage input to AC and then rectify them back to DC, passing through
a high frequency transformer between the second conversion. This topology has a significant volt-
age gain in comparison to the non-isolated and are much better suited to fast-charging stations due
to this [5].
2.4 Topology of Bidirectional DC-DC Converters 9
Figure 2.6: Diagram of bidirectional EV charging stations in grid [5]
Figure 2.7: DC-DC topology flowchart [5]
2.4.1 Isolated Converters
Isolated bidirectional converters are the main option employed for air crafts, renewable energy
sources and electric vehicle charging, its galvanic isolation bring high gain ability as a consequence
of the turn ratio of the windings, making it appropriate for a wide range of input voltage and load
regulation [23]. The isolation will also facilitate the possibility of various input and/or output
topologies and are much better suited to high power application when comparing to non-isolated
converters. It’s isolation also provides a greater safety element for the equipment.
10 State of the Art
The Dual Active Bridge (which will be further explored in Chapter 3) fits in to this category,
and some of the main isolated DC-DC converter topologies that compare to it will be briefly
explored below.
2.4.1.1 Bidirectional Flyback Converter
Also know as the isolated Buck-Boost bidirectional converter, it is implemented, as seen in Figure
2.8, by switching the inductor of a Buck-Boost converter with a transformer for galvanic isolation.
The transformer design is extremely important and a voltage clamp snubber is required for current
leakage protection.
Figure 2.8: Bidirectional Flyback DC-DC Converter [5]
2.4.1.2 Isolated Cuk & Sepic/Zeta Bidirectional Converter
Made to introduce the benefits of magnetic isolation to the original Cuk converter. Figure 2.9
shows the converters layout, providing isolation between both input and output sides with high
voltage gain. A useful feature this converter possesses for renewable energy systems is the cou-
pling of the input and output inductors leads to the elimination of current ripples [24].
Figure 2.9: Isolated Cuk & Sepic/Zeta Bidirectional DC-DC Converter [5]
2.4.1.3 Push-Pull Bidirectional Converter
The circuit in Figure 2.10 was created to allow bidirectional flow on the original push-pull con-
verter; it converts the power using a multi-winding transformer. For high power applications a
three-phase bidirectional push-pull converter is proposed.
Figure 2.10: Push-Pull Bidirectional DC-DC Converter [5]
2.4.1.4 Forward Bidirectional Converter
A bidirectional model of the normal forward converter. This converter achieves zero voltage
switching when a clamped circuit is used. This topology also brings about several hybrid con-
figurations with the other isolated topologies mentioned.
Figure 2.11: Forward Bidirectional DC-DC Converter [5]
2.4.1.5 Dual Half-Bridge Bidirectional Converter
An alternative to the DAB converter (the converter focused on this dissertation and seen more in
depth in the next chapter), the Dual Half-Bridge, Figure 2.12, is used for applications with lower
12 State of the Art
power, having the number of power switches reduced from eight to four, and using two voltage-fed
half-bridges in either sides. This converter may also be configured with a current-fed topology in
the secondary side, which enables continuous current waveform.
Figure 2.12: Dual Half-Bridge Bidirectional DC-DC Converter [5]
2.4.1.6 Half Bridge-Full Bridge Bidirectional Converter
Figure 2.13 shows the Half Bridge-Full Bridge Converter: it is composed by a voltage-fed half-
bridge and full-bridge topology in the primary and secondary sides, respectively. Its lower count
of switches makes for a simpler control in comparison to the DAB. This converter is useful for
UPS topology integration.
Figure 2.13: Half Bridge-Full Bridge Bidirectional DC-DC Converter [5]
2.4.2 Control Schemes for Bidirectional Converters
A big factor to extract the maximum efficiency of a converter is the correct choice of control
schemes, and there are numerous possibilities for bidirectional DC-DC converters. Nevertheless
the main control schemes utilized for bidirectional DC-DC converters in electric vehicle applica-
tions are usually the PI, PID, sliding mode, Fuzzy and Digital control.
The PID and PI control is highly regarded due to its simple implementation, it relies on a
control loop feedback system that receives a reference value set for its desired output, which is
subtracted to the current output to give an error component, this error will be adjusted and tuned
according to the three parts of the PID: Proportional; Integral and Derivative components, for
the PI controller, only the Proportional and Integral parts are used for the calculation. These
control techniques are extremely effective when it comes to the managing the power flow in both
directions. Figure 2.14 shows the PID system diagram.
Figure 2.14: PID control feedback diagram
The PI control will be further explored on PSIM software simulations thought the dissertation.
Document [25] shows a robust design of the PI control, exploring it’s advantages and limitations.
For the sliding mode control, which its feedback diagram can be observed in Figure 2.15,
it’s objective is to alter the dynamics of a non-linear system and to stabilize it with uncertainties
and disturbances. To do so, it defines a sliding surface in the state space, and a control law
is designed to drive the system’s trajectory onto this surface, keeping it there. The control law
includes reaching control, which attracts the system toward the sliding surface, and sliding control,
which drives the system along the surface towards the desired equilibrium point. The key concept
behind sliding mode control lies in the Lyapunov stability theory, which is used to analyze the
stability properties of the sliding surface, ensuring that the system’s state trajectory converges to
the sliding surface and remains on it despite the aforementioned uncertainties and disturbances.
Figure 2.15: Sliding mode control feedback diagram

14 State of the Art
Fuzzy control is a dynamic and adaptive control technique used in DC-DC converters to regu-
late their operation. Unlike traditional control methods, fuzzy control doesn’t rely on precise math-
ematical models. Instead, it uses linguistic variables and rules to handle the converter’s complex
and nonlinear behavior, defining fuzzy sets for input and output variables. Based on these linguistic
variables and rules, the fuzzy controller makes decisions and adapts the control action accordingly.
Fuzzy control effectively handles uncertainties, parameter variations, and disturbances commonly
encountered in DC-DC converters. By employing fuzzy logic, the control action can be fine-tuned
to optimize efficiency, maintain stable power conversion, and ensure reliable performance under
varying operational conditions.
Figure 2.16: Fuzzy control feedback diagram
Document [26] analyses the fuzzy control and sliding mode control on the DAB converter for
battery charge and discharge applications. For the fuzzy control, less measurements and no prior
knowledge are needed for its design. The sliding mode offers a faster converging and more robust
control.
In digital control systems, continuous-time error signals are first converted into a digital format
that can be processed by a computer, using an Analog-to-Digital (A/D) interface. Once the error
signal is in digital form, it undergoes processing within the computer, which is then fed to a
controller, and the controller’s output injected into the system, which in this case is the converter.
This way, the digital controller can efficiently and precisely regulate the system’s behavior based
on the processed error signal, enabling precise and responsive control of the converter’s operation.
The feedback diagram of this control scheme is seen in Figure 2.17.
Figure 2.17: Digital control feedback diagram

For a precise small-signal model of a DAB converter, a thorough understanding of the modu-
lation method and the inclusion of electromagnetic interference (EMI) filters, accounting for their
interactions with the DAB, are essential, as analysed in the article [27]. High-power converters are
increasingly relying on the use of Digital Signal Processors due to their cost-effective high com-
putational performance. Digital implementation offers a greater flexibility compared to analog
electronics, high immunity to electromagnetic interference, and the ability to monitor processes
and faults using external interfaces or network connections. Hence, a digital control platform
was employed to regulate the article’s DAB, and the discrete-time transfer function derived from
discrete modeling, readily utilized for designing the controller.
Table 2.2 shows the benefits and limitations of the four control schemes suggested.
Control Control Benefits Limitations

Scheme Problems
PID -Power flow control -Low cost -Low efficiency
-Minimizing switching time -High reliability -Lack of robustness with un-
between two directions -High dynamic performance certainties and disturbances
-Reducing dead time of -Weakness in avoiding large
switches transient between directions
-Protecting the elements
from over current
Sliding Mode -Considering external per- -Reference tracking -Accurate parameter and
turbations in large signal -Fast and finite-time re- state information are needed
-Dealing with severe varia- sponse
tions in load and line -Robustness against param-
eter variation and external
perturbation
-Ability to characterise the
system under small and
large conditions
Fuzzy -Minimizing power con- -Fast and robust response -Sensitive to expert knowl-
sumption from the grid -Applicability to nonlinear edge
-Attaining smooth perfor- and inaccurate systems
mance of super-capacitor in
charge/discharge
-Minimizing control time
Digital -Accurate small-signal mod- -High EMI immunity -Difficult to implement the
Control elling of DAB -Enhances the efficiency and electronic configuration of
-Changing power flow direc- charge/discharge speed nonlinear control law
tions smoothly with current -Higher flexibility -Requiring great ana-
protection -Fault monitoring log/digital processing effort
-Improving the transient re- -Ease of use
sponse’s speed -Improved reliability
Table 2.2: Summary of control schemes in bidirectional converter [5]
16 State of the Art
Chapter 3
Background on the Dual Active Bridge
The DAB converter is one of the favoured bidirectional isolated converters, it consists of a high
frequency transformer in between two full H-bridges. This converter is gaining traction among
Micro grids and Electric Storage Systems, not only for its bidirectional adjustable power flow ca-
pabilities but also for its simple structure, high power density, wide voltage conversion gain range
and soft switching features [28]. It’s high switch count (eight power switches), accompanying
the galvanic isolation in this topology makes it an adequate fit for high power applications [29]
such as DC fast charge for EVs. For these applications the DAB is employed to regulate the DC
voltage supplied to the vehicle from the three-phase rectifier. In the following sections the theo-
retical concepts of the DAB’s operating system with a PI controller and Single Phase Shift will be
explained.
3.1 Theoretical Development
Figure 3.1 shows a circuit diagram of the DAB, as touched previously, it consists of a high-
frequency transformer and connects two similar H-bridges made up with transistors, the first one
being a DC-AC inverter and the second being a AC-DC inverter, together making a full DC-DC
converter. It allows the control of power flow in the circuit depending on the phase shift between
the square waves (Vp and Vs ) applied to the bridges (Single Phase Shift), and will affect the MOS-
FET or transistor pair that is conducting simultaneously making a total of 4 different operating
modes. Observing the waves on Figure 3.2 and the Table 3.1 along with Figure 3.3, we can verify
the change of the current wave in the coil (iL ) at each mode of operation and what each operating
mode means in regards to the conducting components.
The difference applied in the angle φ , will affect the waveform current in the coil and, conse-
quently, the average current in the output which will be a rectified (iL ) wave. These changes will
impact the output power to the battery.
When working with the Dual Active Bridge, the maximum power flow is obtained when the
offset angle φ between the waves is equal to π/2 rad as it obtains the highest possible average
current. When the secondary wave is lagging behind the primary, a flow of current will be directed
17
18 Background on the Dual Active Bridge
Figure 3.1: DAB converter circuit [6]
Figure 3.2: DAB waveforms for forward power flow: Primary Voltage, Secondary Voltage and
Coil Current respectively [6]
Table 3.1: Operating modes of the DAB
Operating Modes Transistor Pairs Conducting

1 S11, S14 & S22, S23
2 S11, S14 & S21, S24
1’ S12, S13 & S21, S24
2’ S12, S13 & S22, S23
3.2 Design 19
Figure 3.3: DAB Current flow at each operating mode.
from the primary to the secondary side, and conversely, when the primary wave is lagging behind
the secondary, the direction of the current flow will be reversed, bringing about its bidirectional
feature. This versatility in power flow direction is one of the key advantages of the DAB converter
in V2G operations.
3.2 Design
The design of the DAB is critical for its performance, and requires careful considerations of vari-
ous factors.
The DAB operates at high frequencies, enabling smaller and more efficient components, one
of which is the transformer. For the transformer design, it is extremely important to decide its
turn ratio, n = Vs /Vp . There is a possibility of adjusting the wave to a boost (Vp > Vs /n), buck
(Vp < Vs /n) (both of them can be seen in the iL wave on Figure 3.2), or ideally equalling Vs /n
to Vp . This ideal scenario would help the efficiency of the DAB by reducing the rms current and
reactive power.
During the DAB design it is also important to consider its peak power operating conditions.
As addressed previously, the current, and consequently the power, would be adjusted by the angle
φ , ranging on the values [-π/2, π/2], however, despite the maximum possible power flow being
at these two points, its range should be decreased, as the power flow flips at these angles, from
positive flow to negative, and to allow room for transients. At maximum power, the maximum
angle should be close to, but not equal to π/2.
Regarding the size of the coil, it is important to inspect the equations for a Dual Active Bridge,
which, with a deduction of the area within the output current wave we will be able to arrive at the
equations 3.1 and 3.2 for the output current and power respectively.
Vi φ (1 − φ /π)
Io = (3.1)
2nπFs L
VoVi φ (1 − φ /π)
Po = Vo Io = (3.2)
2nπFs L
With these equations we can verify that the size of the coil is essential to maximise the power
extraction, as the bigger the coil, the smaller the maximum power. By modifying the expression
3.2 as a function of the coil (L), we can calculate its size for the power needed.
3.3 Control
For the circuit control, it is important to decide which control scheme to use and how to handle
the transistor switching. There are many different variations that can be opted for both these
operations, with each having positive and negative characteristics.
For the choice of control scheme the variations explored in Chapter 2.4.2 for bidirectional DC-
DC converters are all suitable for DAB applications, with the PI controller being chosen to further
explore the DAB with PSIM software testing.
For the PI controller, its main objective is to control the current by establishing a reference
value for the power wanted, and comparing it to the exact value measured on the output, and
altering the angle of phase shift φ according to its proportional and integral components. Figure
3.4 shows the working diagram theory for a PI controller working on the DAB converter.
Figure 3.4: PI controller diagram for DAB applications
As for the switching strategy, the aim is to maximize power range and optimize ZVS char-
acteristics. There are four usual switching strategies for the DAB: Extended phase shift (EPS),
Double phase shift (DPS), Triple phase shift (TPS) and the one most prominent in this thesis the
Single phase shift (SPS) [7].
Single Phase Shift concept is the more traditional route when it comes to controlling the DAB,
and has already been touched on in the theoretical development section. In comparison to the SPS
which relies on one phase difference between the primary and secondary sides, the EPS control
adds an equal inward shift between the primary and the secondary side on the basis of SPS and
3.4 Interleaved Dual Active Bridge 21
the DPS adds an inward shift in both at the same time [7], their respective waves can be seen in
Figure 3.5 and can be compared to the single phase shift waves in Figure 3.2.
3.4 Interleaved Dual Active Bridge

The Interleaved DAB converter (which can be seen simulated in Figure 4.2) is a sophisticated
power electronics topology that combines the benefits of both the Dual Active Bridge and inter-
leaving techniques, and has its main ability in efficiently managing high-power conversions. It
consists of multiple identical dual active bridge converters connected in parallel.
The purpose of interleaving is to improve the overall performance and efficiency of the system.
By splitting the power among multiple parallel-operated bridges, the interleaved dual active bridge
converter reduces the current and voltage stress on each converter, resulting in various advantages,
for instance: reduced device stress, as it allows the distribution of power across several power
semiconductor devices, leading to lower current and voltage ratings on each device, extending the
converter’s lifespan and increasing reliability; reduced switching losses, as it reduces the switching
frequency seen by each bridge, thereby lowering switching losses; reduced output ripple, among
other benefits [30].
This variation of the DAB can mitigate the challenges associated with high currents and volt-
ages, leading to improved efficiency, reduced stress on components, and enhanced power handling
capabilities.
3.5 Conclusions
This chapter presented the theory behind the workings of the DAB converter, how it functions
and it’s waveforms. It was understood it’s design and important parameters to set to obtain ex-
pected results. This section demonstrated the DAB converters flexibility in regards to adjusting
power flow in not only one direction but in both, and it’s great response to different power levels,
exhibiting it’s amazing potential for EV fast charging station.
Figure 3.5: Waveforms for (A) EPS and (B) DPS [7]
Chapter 4
Simulation Test
To confirm the validity of the theoretical research, a set of simulations werw conducted using the
PSIM software. PSIM (Power System Simulation for Industrial Motors) is a simulation software
used to analyse and design electrical power systems. It is commonly used in the fields of electrical
engineering and power electronics to model, simulate, and test power system components. The aim
is to simulate a Dual Active Bridge converter working in a DC fast charger for electric vehicles,
observing the behaviour of the power system under different operating conditions and test its
performance before implementing a scaled model of the DAB in the experimental phase.
4.1 Objective
The objective of this simulation was to emulate a Dual Active Bridge converter working in a 100
kW DC fast charging station, charging a Volkswagen ID.4 battery system according to its fast
charging curve in Figure 2.5, and subsequently analysing and interpreting appropriate waveforms
to validate its correct implementation. Further simulations were made with an interleaved DAB to
highlight its additional properties.
4.2 Specifications and Circuit Setup
The simulation was set up according to the specific requirements of EV charging in real-life situ-
ations, and the DAB components were designed based on the previously studied theory. Table 4.1
shows the specifications made for the simulation.
Table 4.1: Simulation value table.
23
24 Simulation Test
Vi 650 V
Vomin 300 V
Vomax 400 V
Vo80%SoC 380 V
n 0.5
Pomax 100 kW
φ −π/6, 2π/5
L 26.5µH
fs 20 kHz
The voltage Vi entering the DAB was chosen from a realistic value for an EV charger, it
simulates the DC voltage arriving from the AC-DC converter that is connected to the public 3-
phase grid, and is usually around 650 V for a direct connection to the grid.
Voltage on the battery was considered on the basis of the battery specifications of the VW ID.4
model [4]. It was calculated by multiplying the voltage value of a fully charged lithium cell (4.2
V) by 96 the number of cells in series in the vehicles battery, this would give 400 V maximum
output voltage to the battery. It’s model was based on a Thevenin circuit equivalent model for an
EV battery [31].
The switching frequency was predetermined to be 20 kHz, as the DAB functions with high
frequencies, and finally φ was chosen to be a little below its maximum possible quantity for their
to be a safe distance in angle before a full power direction switch occurs, the choice of maximum
φ value was made by subtracting 0.1π from the 0.5π that set the maximum power flow. The
bidirectional capability angle φ was limited to a minimum of negative 30 degrees.
The next components were calculated with the theory gathered. Value n for the turn ratio in
the transformer uses the known equation n = Vo /Vi , Vi and Vo depends on the state of charge of the
battery, so to minimize the reactive power, Vo was set at approximately the middle point between
maximum charging voltage (with car at 80% SoC) and minimum Vo . The coil was chosen by
adapting the equation 3.2 and substituting all the other known variables.
To simulate the DAB in conditions closer to real life, IGBT values where chosen according
to suitable transistors on the market. The respective values to fill would be the saturation volt-
age, transistor resistance, diode forward voltage and diode resistance. After analysing several
IGBT datasheets, the model chosen for the transistors in the first bridge would be the Infineon
FF300R12KS4 model. This transistor has a nominal voltage capacity of up until 1200 V and
nominal current until 300 A, which is within the limitations for the first bridge. By looking at
the datasheet [32], we can find it’s saturation voltage of 3.2 V and diode forward voltage of 2 V,
to calculate the transistor and diode resistance we would need to find the value for the maximum
current rating in the datasheet which would be at 400 A, by dividing the saturation voltage by
the maximum current rating we would obtain the transistor resistance (0,008 Ω) and by dividing
the diode resistance by the same maximum current rating we obtain the diode resistance (0.005
Ω). For the second bridge the voltage would be halved due to the transformer, however higher
4.2 Specifications and Circuit Setup 25
current levels would pass through the transistors. The IGBT model Infineon FF400R06KE3 was
chosen, with a nominal voltage of 400 V and a nominal current of 600 A, it would be well within
the maximum power output possible. By analysing it’s datasheet [33], and using the same process
carried out for the first IGBT model the values found where: Saturation Voltage = 1.45 V; Diode
Forward Voltage = 1.55 V; Maximum current rating = 500 A; Transistor Resistance = 0.0029 Ω
and Diode Resistance = 0.0031 Ω.
With these specifications the circuit seen in figure 4.1 was built, and for further testing Figure
4.2 shows the circuit for the interleaved DAB.
Figure 4.1: DAB circuit on PSIM.
Figure 4.2: Interleaved DAB circuit on PSIM.

26 Simulation Test
4.3 Control
4.3.1 Phase Shift Control Implementation
On figure 4.3 it can be observed the control signal used for the Single Phase Shift of the DAB,
implemented to the IGBT pairing. A sawtooth wave was used in the primary and secondary bridge
pairings, with the second bridge having a reference angle (phi) in degrees, multiplied by -1 that
would determine the phase shift. The respective outputs would then pass through a sine compo-
nent, which produces two sine waves shifted by the reference angle set, and when comparing it
with the neutral creates a PWM control with the shift applied on the second bridge. As discussed
earlier this shift is crucial for the DAB converter.
Figure 4.3: Phase shift control Circuit on PSIM.
Figure 4.4 shows the step by step waveform transformation until the final phase shifted waves.
1) Sawtooth and Sawtooth with shift; 2) Shifted sine waves; 3) Final phase shifted waves;
The interleaved phase shift control was carried out as per the theory suggested, similarly to
the simple DAB, the square waves for the first DAB component of the interleaved model where
formed with a sawtooth and a sine comparison to the neutral with the secondary bridge being
shifted "phi" degrees. For the second DAB component, the whole system was lagged 90 degrees
before building the square waves. Figure 4.5 shows the phase shift control constructed, and Figure
4.6 shows the square waves with a phase shift of 30 degrees; Note that on the second graph of this
figure the waves are lagging 90 degrees in comparison to the first.
4.3.2 PI controller
To establish the PI controller, based on the one seen on Figure 4.4, first the charging curve for
the desired power output (P_ch) was built in a code block, and was divided by the filtered Voltage
output of the converter, to obtain the reference current. To this reference, the filtered output current
4.3 Control 27
Figure 4.4: Phase shifted PWM wave built step by step.
Figure 4.5: Phase shift control circuit for interleaved DAB on PSIM.
would be subtracted to receive the error value of the system. The PI function block would then
transform the error in to the angle phi which would be inserted as the phase shift on the PWM
transistor waves.
The charging curve was loosely based around the one seen on Figure 2.5. Its functioning plan
28 Simulation Test
Figure 4.6: 30 degree phase shifted waves for first and second DAB.
Figure 4.7: PI Controller Circuit on PSIM.
was that a power 100 kW was applied to the battery for the first seconds, constantly decreasing to
65 kW and maintaining it, the curve final seconds would vary depending on the test. Figure 4.8
and 4.9 shows a power charging curve example.
4.4 Results
4.4.1 30º vs 72º Phase Shift at Permanent Regime
Several simulations where carried out to observe the waveforms of the DAB in different condi-
tions. The first run of simulations was to analyse the waveforms of the voltage in the primary and
4.4 Results 29
Figure 4.8: C block code for power charging curve.
Figure 4.9: PSIM power charging curve graph.
secondary windings, as well as the voltage and current in the coil and its subsequent output current
at different levels of phase shift.
Figure 4.10 and 4.11 show these waves with the converter performing at φ equalling 30º and
72º respectively. By analyzing the graphs disposed we can clearly see a larger shift implemented
in the secondary square waves of figure 4.10a in comparison to figure 4.11a. This shift can also
be seen in the voltage of the primary and secondary windings of the transformer. At the primary
winding the square wave’s peak voltage reaches 650 V, the Vi value, and after passing the 2:1
transformer it is reduced to 325 V in the second winding. These waves are shifted according to
its introduced φ angle. Figure 4.10c and 4.11c, demonstrate the voltage and current in the coil.
VL reaches a maximum value of Vi +Vo /n for the time that the primary wave leads the secondary
(operating mode 1 of Table 3.1), and gets to −(Vi + Vo /n) at operating mode 1’ of Table 3.1.
Operating modes 2 and 2’ will give values of Vi − Vo /n and −Vi + Vo /n, respectively. In the
example these almost reach zero as the transformer ratio almost equals Vi to Vo /n. VL and IL show
the AC component of the DAB.
Io demonstrates the DC rectification of the AC component of the DAB. As the angle φ in-
creases, so does the the average of the current output. By comparing Figures 4.10d and 4.11d,
30 Simulation Test
with 72 degree angle phase shift the current output maximum value increases, and so does it’s
average value. For φ =72º the power output reaches the maximum desired power of 100 kW while
at φ =30º the power reaches around 58 kW.
4.4.2 Reverse Flow in Permanent Regime
To observe the bidirectionality of the DAB, the same waves on the previous subsection were anal-
ysed in permanent regime with an angle of φ =-30º. In Figure 4.12a and 4.12b the secondary wave
leads the primary one, and as the theory suggests, this creates a flow towards the reverse direction.
VL and IL follow the same principle as the previous explained waves. The main difference comes
in the output current waves, at an angle φ of -30º the Io wave is reversed in comparison to the
original power flow. The average current in this graph obtains a negative value, which will thus
give way to a reverse power flow in the converter.
4.4.3 PI Control Response
To inspect the transitory response of the converter, two reference power charging curves where
coded, both had the same curve for the first 4 seconds, and after that the first reference in Figure
4.13a abruptly ends, while the second in Figure 4.13b constantly decreases until φ =-20º.
In these figures we can see in red the wanted power curve and in blue the power flow with
PI response to follow the reference, in the graph below each figure there is the phi response that
changes accordingly to obtain the power needed, with a maximum of 72º and a minimum in figure
4.13b of -8º.
The graphs in Figure 4.14 show an overview on output current against reference current. It is
evident that the waveform of Io decreases in size as the reference current decreases in value. This
decrease in peak to peak value diminishes the average current flowing and thus the power adjusts
to a lower level.
4.4.4 Step responses
The final simulations made had the objective to analyse the dynamic response of the output cur-
rent’s to a step in the reference current.
Figure 4.15a shows the response to a step to 0 A, through the figure after the suuden decrease
in the reference current, the average current Io decreases rapidly, and reaching an average constant
value of 0 A in 1.25ms. Figure 4.15b shows a step to a negative value, Io reaches a constant value
in quicker fashion in 0.2ms.
4.4.5 Interleaved DAB
Further simulations were carried out, to also analyse the 2 level interleaved DAB, seen in Figure
4.2.
4.5 Conclusions 31
Figure 4.16a, shows the 20 kHz frequency square waves applied to the converter and, as seen
in the theory development, the second PWM signal would be shifted 90º from the first. This same
shift can be seen in the winding voltage (Figure 4.16b).
In Figure 4.16, several readings were made to the converter for analysis. In the first graph
screen, the current at the output of the first bridge (Idc1) and second bridge (Idc2) as well as
the current after the bridges (Ibat1) are shown. It is confirmed that the effective frequency at the
output of the bridges are twice the switching frequency that the interleaved DAB operates, which
propagates effective sharing of the load. Ibat2 and Vo1 show the values of current and voltage at
the battery load simulator, this would equate these values to 115 kW.
4.5 Conclusions
With these simulations, the theory was put into practice by implementing the DAB converter in
a real life DC charging station simulation. The converter was designed, attending to real life
values, and the PI controller was developed and showed impressive responses, corroborating with
the fluidness of the phase shift seen in Chapter 3, by following a charging curve with impressive
reactions in regards to it’s power output changes.
32 Simulation Test
(a) Square waves with 30 degree phase shift.
(b) Voltage in the primary and secondary windings of the transformer with 30 degree phase shift.
(c) Voltage and current on the coil with 30 degree phase shift.
Figure 4.10: Waveforms for 30 degree phase shift.

4.5 Conclusions 33
(d) Current output with 30 degree phase shift
Figure 4.10: Waveforms for 30 degree phase shift.

34 Simulation Test
(a) Square waves with 72 degree phase shift.
(b) Voltage in the primary and secondary windings of transformer with 72 degree phase shift
(c) Voltage and current on the coil with 72 degree phase shift
Figure 4.11: Waveforms for 72 degree phase shift

4.5 Conclusions 35
(d) Current output with 72 degree phase shift
Figure 4.11: Waveforms for 72 degree phase shift

36 Simulation Test
(a) Control signals: square waves with a negative 30 degree phase shift
(b) Voltage in the primary and secondary windings of transformer with a negative 30 degree
phase shift
(c) Voltage and current on the coil with a negative 30 degree phase shift
Figure 4.12: Waveforms for a negative 30 degree phase shift.

4.5 Conclusions 37
(d) Current output with a negative 30 degree phase shift
Figure 4.12: Waveforms for a negative 30 degree phase shift.
(a) Charging power and angle phi response with abrupt end.
(b) Charging power and angle phi response with power flow switch
Figure 4.13: Waveforms for the charging power and angle phi PI response.
38 Simulation Test
(a) Overview of output current with reference power to 0 kW
(b) Overview of output current with reference power to -20 kW
Figure 4.14: Overview of output current against reference current.

4.5 Conclusions 39
(a) Zoomed output current step response to 0 kW
(b) Zoomed output current step response to -20 kW
Figure 4.15: Zoomed waveforms for the output current step response.
40 Simulation Test
(a) Square waves with a 30 degree phase shift
(b) Voltage in the primary and secondary windings of both transformers with a 30 degree phase shift
(c) Voltage and current on the coil with a negative 30 degree phase shift
Figure 4.16: Waveforms for interleaved DAB with 30 degree shift.

Chapter 5
Experimental Test
After the simulation process, experimental work was started to test the DAB in scaled real condi-
tions, with hardware. For this tests to work out the microcontroller must be chosen and the square
waves implemented, as well as the values of voltage input picked, as well as the transformer and
coil calculated.
It was established that the DAB would work at a frequency of 20 kHz and would receive a DC
input of up to 60 V, a value sacaled by 10 compared to the PSIM simulation.
5.1 Microcontroller
The choice of microcontroller, that would be used to control the waves, fell to the STM32F429
[5.1]. It is a microcontroller from the STM32F4 series, which is a family of based on the ARM
Cortex-M4F architecture. The STM32F429 is typically used in applications that require high per-
formance and a wide range of connectivity options, such as in medical devices, industrial control
systems, and consumer electronics. This controller might be over the top for the DAB control
proposed. However, due to having slight contact with it in past classes, interest was gained in
evolving the knowledge in this realm. This model is based on a nucleus of 32bits and can work at
frequencies of up to 216 MHz. It offers 13 timers and its pins at "HIGH" value are limited to 3.6
V.
5.1.1 Square Wave Implementation
To control the DAB, as the theory suggests, 4 waves must be implemented, 2 of them comple-
mentary to one another and with phase shift, when compared to the first. For this, the following
configurations must be made:
1. Clock configuration;
2. Timer configuration;
3. Dead Time insertion;
41
42 Experimental Test
Figure 5.1: STM32F429 Development Board.
4. Signal declaration on the firmware.
5.1.1.1 Clock Configuration
On the clock configuration interface, in the STM32CubeIDE software, a value of 25MHz was set
for the system clock.
5.1.1.2 Timer Configuration
For the timer configuration, the idea was to set up 3 timers, each with 1 set of complementary
waves. These timers would be triggered by the previous timer, to allow for the phase shift between
complementary pairs. Figure 5.2, shows the timer interface.
To obtain the PWM waves, the "PWM Generation channel" was chosen at each timer, and for
this wave to work at 20 kHz calculations were made using the following equations:
APB.T IM.CLOCK
T IM.CLOCK = (5.1)
PRESCALAR
T IM.CLOCK
f requency = (5.2)
ARR
CCR
DutyCycle(%) = 100 (5.3)
ARR
APB.TIM.CLOCK is the system clock that was set to 25 MHz, and by knowing the desired fre-
quency of 20 kHz, the calculation lead to the values that can be seen on Figure 5.2: PRESCALER
set to 5 and AutoReloadRegister (ARR) set to 250. The values were subtracted by 1 as in the
STM32 the starting bit is 0.
5.1 Microcontroller 43
Figure 5.2: STMCubeIDE timer interface.
After setting up the frequency of the wave, the PWM wave must be set up. For this project the
PWM was generated in centre align mode. Figure 5.3 shows the main function of the center align
mode, it allows for the PWM to be built in the center permitting easier control of the phase shift if
bidirectional flow is applied. A comparison between the center aligned and usually standard edge
aligned mode can be observed in Fig.5.4.
Figure 5.3: Centre align mode principle, [8].
The trigger control is an essential component for the phase shift to work was the configuration
Figure 5.4: Centre align mode comparison to edge aligned, [9].
of the triggers and for each timer to work as a master or slave. Table 5.1 shows the internal trigger
connections between timers.
Table 5.1: Internal Trigger Connections
Slave TIM ITR0 ITR1 ITR2 ITR3

TIM2 TIM1 TIM8 TIM3 TIM4
For the project in hand, the three timers initialized were: TIM2, TIM3 and TIM4. TIM 2 would
act as the PWM in the primary bridge, TIM 4 would act as the PWM for the secondary bridge,
and TIM3 acts as a trigger intermediary to apply the phase shift between TIM2 and TIM4. For
this, both TIM3 and TIM4 work in slave mode activated by the signal ITR1 and ITR2 respectively,
meaning TIM3 is a slave of TIM2 and TIM4 a slave of TIM3. To control the angle φ , an ouput
compare channel (OCC) was inserted in the master timers (TIM2 and TIM3), these channels would
activate the PWM signal of their respective slaves when they changed from LOW to HIGH and
the timing of this switch could be controlled by the pulse of the OCC, thus controlling the phase
shift.
Figure 5.5 has an example of the implemented timers and output compares. Timer 2 has two
PWM generators complementing each other and an output compare that sets to one after the pulse
value time has passed, counting from the beginning of the timer initialization. This switch will
activate it’s slave, which in this case is TIMER 3 and the same will happen to the slave/master
dynamic between TIMER3 and TIMER4.
5.2 Transistor Control 45
Figure 5.5: Output Compare with active level on match.
5.1.1.3 Dead Time Insertion
An extremely important factor for the safe behavior of the converter is the insertion of dead time
in the PWM square waves applied to the transistors. This will allow for the soft switching of the
ZVS bridges of the DAB. The dead time is the time that must exist between the commutations in
a leg, and must be high enough so that that the ON/OFF switching does not occur simultaneously.
For the waves at a 20 kHz frequency, a full period will take 50 us, for this time period a dead
time of 1us was chosen which equates to about 2% of the full period. It was implemented by
altering the duty cycle of the complementary waves, and as they were in centre align mode, by
having the duty cycle HIGH proportionally lower to when its LOW, an adjustable dead time was
created.
5.1.1.4 Signal Declaration on the Firmware
After the timer initialization, it is necessary to declare the signals. For this objective, the HAL
library was used in the STMCubeIDE.
Figure 5.6 shows the code that initializes the timers and GPIOs, then it writes the value of the
phase shift by chasing the prescale value of the OC on channel 3 of TIMER3. After that it declares
all the PWM’s and OC’s generated.
5.2 Transistor Control

The transistor control system is isolated by optocouplers, which are used to protect control cir-
cuits, normally with low voltage/current, from more dangerous circuits involving high voltages
and currents. For the correct operation of these optocouplers, it was necessary for the signal to
Figure 5.6: HAL library code for signal declaration
transmit 5V with a minimum current of 10 mA. The microcontroller STM32F429 signals were of
voltage and current lower than necessary, and so, a buffer, the CD4050 was used [34]. This buffer
allows the placing of several input signals to the same level voltage, which will be the buffer’s
supply voltage, in this case the PWM signals will go from 3.6 V to 5 V. To increase their current,
each of these signals was connected to 3 buffers in parallel.
Figure 5.7: CD4050B CMOS Hex Inverting Buffer
5.3 Circuit Build and Setup

For the DAB circuit build the hardware utilized included:
• DC Voltage Source;
• 2x 1ph inverter driver (H-bridge) (Figure 5.8);
• Medium frequency transformer;

5.3 Circuit Build and Setup 47
• Coil;
• DC electronic load.
The DC Voltage source used for the input voltage of the converter would be able to reach up
to 60 V and would connect to the first 1ph inverter driver with a SK25GH063 module [35]. The
second 1ph inverter driver was connected to the first after the coil and transformer series circuit.
The DC electronic load used was the BK Precision 8500 that could receive a maximum input
power of 300 W [36].
A 100 uH coil was welded in series with the transformer, as shown in Figure 5.9. This trans-
former was built with a turns ratio of 40:22, which, at a maximum primary voltage of 60 V the
secondary voltage would decrease to around 33 V, a ratio of 1:0.55. The ferrite used was EE42,
which was an appropriate option for the power and frequency of the experimental project [37]
[38].
Figure 5.8: Single phase inverter with current sensors.
To control the switching of the experimental prototype and input the PWM signals defined
by the microcontroller and enhanced by parallel buffers, a 14 pin connector makes the coupling
between the signals and the 1ph inverter. Figure 5.10a shows the circuit layout of the 1ph inverter
and Figure 5.10b shows the layout of the 14 pin connector.
After pre-testing all the hardware and joining them as explained earlier, the full ecosystem for
the DAB to work was finished and ready for testing. Figure 5.11 shows the full DAB converter
circuit built, with isolated voltage and current probes and test equipment.
Figure 5.9: Coil and transformer series circuit.
(a) Single phase inverter, SK25GH063 module layout
(b) Connector pinouts
Figure 5.10: Connector pins and bridge circuit diagram.

5.4 Results 49
Figure 5.11: Full DAB circuit build.
5.4 Results
5.4.1 Preliminary Testing
Before the functional test at maximum power, preliminary testing is needed to guarantee the cor-
rect operation of the hardware. The first requirement is the right PWM signal with sufficient dead
time produced by the microcontroller as described previously. Figures 5.12 shows the different
conditions surpassed in the PWM signal implementation. 5.12a shows the complementary signal
that is applied to the first bridge with a 1us dead time, which is confirmed in Figure 5.12b, with
the zoomed signal at 1us per square in the oscilloscope.
Figure 5.12c shows the attainable, variable phase shift between the PWM waves, in this case
with an angle φ of 30º. For the bidirectional characteristic the following Figure 5.12d, shows the
phase shift with φ at -30º.
To test the bridge modules, previous testing was also made by applying the PWM signals to
only one of the bridge while the other would act as a diode bridge. Both bridges were confirmed
to be operating without problems and figure 5.13 shows the results on the oscilloscope of this final
preliminary test.
(a) Complementary PWM waves with dead time
(b) 1us dead time
Figure 5.12: PWM signal confirmation.

5.4 Results 51
(c) PWM at φ =30º
(d) PWM at φ =-30º
Figure 5.12: PWM signal confirmation.

Figure 5.13: Full bridge and diode bridge test result. Yellow: PWM; magenta and light blue:
primary and secondary voltages; dark blue: primary AC current.
5.5 Conclusions 53
5.4.2 DAB testing

After corroborating the hardware components and signals, the DAB could be finally tested, to
analyse the waveforms of the primary and secondary windings, as well as the current IL at angle
phase shift of φ = 30º, 0º and 60º.
The first analysis was to observe the waves at an angle of 30º, with the maximum set voltage
value of the load of 24 V. Figures 5.14, shows the electric load values (5.14a) and the oscilloscope
waves (5.14b). The waves presented show in yellow and light blue, the voltage in the primary
winding (Vp in figure 3.1), before the coil, and the voltage in the secondary winding (Vs in figure
3.1), respectively. These measurements shows the difference of 30º in angle between both bridges,
and their peak value represent the input Voltage (60 V) in the primary winding and output voltage
in the secondary (24 V). In pink, the IL shows the AC component, the electric load receives 3 A,
which will equate to an output power of 72 W.
When increasing the electric load voltage from 24 V to 36 V, the respective waves (Figure
5.15b) change in comparison to Figure 5.14b. The peak voltage in the secondary winding will
increase, turning IL from boost to buck, as n = 0.55 and with the new value of load voltage, 36
V/0.55 = 65 V, slightly higher that 60 V. This environment will operate with an output power of
94.3 W.
The second test would alter the angle φ to 60 degrees and revert the electric load voltage back
to 24 V. The Vp and Vs relation (Figure 5.16b) would thus be similar to Figure 5.14b, however with
larger phase shift, which will increase the power output as it was observed in Chapter 2. Current
IL has a bigger area when comparing to the previous examples, the AC component will lead to a
bigger average current to the output, and as seen in Figure 5.16a, the highest current value was
recorded, 4.43 A which will equate to 106.32 W.
The final test would be at φ =0, as could be noted by the overlapping Vp and Vs waves in 5.17b.
The current going to the load equals 0.58 A, not a null value and the IL still has a small AC wave.
This is due to the reactive power still flowing and possibly the PWM signal phase shift may not
exactly be equal to 0 (it should be noted that the tests were made in open loop).
5.5 Conclusions
This chapter presented the experimental phase of the dissertation, where further knowledge on the
practical applications of the DAB where observed. During this process, a thorough implementa-
tion of the PWM signal waves was made with the use of the microcontroller STM32F429, with
careful considerations on the voltage levels applied to the transistors and crucially on the dead
time insertion to allow ZVS. In this portion of the dissertation more problems were faced in com-
parison to the PSIM simulation due to its nature of testing and confirming the right behaviour of
the components to check for any flaws.
(a) Electric load values (Voltage and current input) with maximum 24 V
set.
(b) Oscilloscope waves
Figure 5.14: Testing results for DAB at 30º angle phase shift and load voltage of 24 V.
5.5 Conclusions 55
set.
set.
5.5 Conclusions 57
set.
Chapter 6
Conclusions and Future Work
6.1 Conclusions
As seen throughout this document, it is evident that the electric mobility market is increasing sub-
stantially and with this growth, EV charging stations must be up to par with the growing demand.
The DAB converter will certainly be essential to this growth due to the many advantages explained
in this thesis. This study delved into the theory of this power converter and its contextualization in
regards to EV charging.
Several PSIM simulations were made where the performance of the DAB was evaluated in
different conditions. The essential waveforms for its proper operation was examined and further
analysis in terms of varying phase shift, not only with static evaluations, but also the dynamic
alterations with its control tactic implemented was made to observe the impact on the power output
given to the vehicle’s battery. Among with these tests, its important bidirectional capabilities,
modular construction and the interleaved DAB were closely checked, to verify its V2G capabilities
and to understand the many complexities and the versatility of the converter at hand.
Furthermore, practical experiments of the DAB were put into place to observe and evaluate
the converter without the ideal conditions of a virtual simulation, simulating a scaled environment
where the DAB converter would operate inside a DC EV charger. From the choice of the compo-
nents and to the decision on the testing values, these experiments were executed with careful step
by step testing on the individual components to guarantee safe conditions for the desired results.
Through the experiment the essential DAB waveforms were obtained and the impact that the phase
shift inserted by the microcontroller on the H-bridges had on a load was examined, corroborating
on the theory presented in previous chapters.
In conclusion, the DAB was extremely impressive due to its power range and fluid power
flow. Its efficiency can be optimized with a more robust controller design and with minimization
of switching losses, and with its optimized abilities it is a perfect suite for a DC fast charging
station, either from its fluid power output control to protect the car battery or because of it’s
simple integration with V2G operations that may arise in the future.
59
60 Conclusions and Future Work
Power electronics are crucial for EV evolution and this converter may in the future be one of
the main proponents of electric mobility throughout the world.
6.2 Future Work

After the completion of the work detailed in the document, additional testing conditions were tried
for a more complete understanding of the DAB in real conditions. More specifically a transient
analysis of the converter. In this trial, the phase shift would change to evaluate in the oscilloscope
the instant change in current value when φ change from 0 to 30 degrees. However, due to the
microcontroller’s output compare reset time, the PWM wave’s frequency inserted to the converter
would change at the instant of the angle shift, thus compromising the integrity of the converter and
validity of the results obtained.
With more time to gain experience with the microcontroller chosen, or by altering the choice
made, there would be several future testing that could be made to enrich this work, all relying
on altering conditions in a continuous test. The instant of direct change of angle would be an
interesting experiment to run, and further down the line a full control strategy, which in this case
would be the PI control, could be tested to analyse its efficiency and real life performance. With a
real battery used as the load, testing the bidirectional V2G component would be an incredible test
to run for a complete, full DAB for EV charging at scaled condition testing.
References
[1] International Energy Agency. World Energy Data, 2022. URL: https://www.
worldenergydata.org/world-electricity-generation/.
[2] H. Ritchie, M. Roser, and P. Rosado. CO2 and Greenhouse Gas Emis-
sions. Our World in Data, 2020. URL: https://ourworldindata.org/
co2-and-greenhouse-gas-emissions.
[3] go-e GmbH. Ac-dc charging. https://go-e.com/en/magazine/ac-dc-charging.
[4] EV Database. Volkswagen ID.4 Pro. https://ev-database.org/car/1627/

Volkswagen-ID4-Pro.
[5] S. A. Gorji, H. G. Sahebi, M. Ektesabi, and A. B. Rad. Topologies and control schemes
of bidirectional DC–DC power converters: An overview. IEEE Access, 7:123994–124017,
August 2019. doi:10.1109/ACCESS.2019.2937239.
[6] J. Nan. A comparative study of three-phase dual active bridge and single-phase single-stage
converters for EV charging applications. Arizona State University, 2011.
[7] J. Zou and Y. Sheng. Research on control technology of dual active full bridge DC-DC
converter. Journal of Physics: Conference Series, 2030:012011, 2021. doi:10.1088/
1742-6596/2030/1/012011.
[8] STM32F303x6 center-aligned PWM ADC timing HAL, 2021.

URL: https://stackoverflow.com/questions/46241956/
stm32f303x6-center-aligned-pwm-adc-timing-hal.
[9] How to view center-aligned PWM on an oscilloscope, 2021. URL:

https://electronics.stackexchange.com/questions/418941/
how-to-view-center-aligned-pwm-on-an-oscilloscope.
[10] BP. BP Statistical Review of World Energy 2021, 2021. URL:

https://www.bp.com/content/dam/bp/business-sites/en/
global/corporate/pdfs/energyeconomics/statistical-review/
bp-stats-review-2022-full-report.pdf.
[11] Intergovernmental Panel on Climate Change. Special report on renewable energy

sources and climate change mitigation, 2011. URL: https://www.ipcc.ch/report/
renewable-energy-sources-and-climate-change-mitigation/.
[12] European Environment Agency. Report on Transport and Environment, 2018.

URL: https://www.transportenvironment.org/annual-report-2018/
documents/TE-Annual-Report-2018.pdf.
61
62 REFERENCES
[13] Ministério do Ambiente e da Ação Climática. Roteiro para a neutralidade car-

bónica 2050: Roteiro da estratégia nacional de longo prazo 2020-2050 [roadmap
for carbon neutrality 2050: Roadmap for the national long-term strategy 2020-
2050]. https://www.portugal.gov.pt/download-ficheiros/ficheiro.
aspx?v=5e5cf276-ef0c-4aa5-bd1c-9c9af23979b5, 2020.
[14] L. Friedman. Biden’s electric vehicle goal: 50 percent of sales by 2030. The New
York Times, 2021. URL: https://www.nytimes.com/2021/08/05/briefing/
electric-vehicles-president-biden-climate.html.
[15] C. Argue. EV battery health: How to extend electric vehicle battery life, 2021. URL:
https://www.geotab.com/blog/ev-battery-health/.
[16] J. Smith. The different levels of EV charging. Forbes, 2022. URL: https://www.
forbes.com/wheels/advice/ev-charging-levels/.
[17] Power Sonic Corporation. Levels of EV charging: An overview, 2021. URL: https:
//www.power-sonic.com/blog/levels-of-ev-charging/.
[18] W. Kempton and S. E. Letendre. Electric vehicles as a new power source for electric utilities.
Transportation Research Part D: Transport and Environment, 2(3):157–175, 1997. doi:
10.1016/S1361-9209(97)00001-1.
[19] D. Lauinger, F. Vuille, and D. Kuhn. A review of the state of research on vehicle-to-grid
(V2G): Progress and barriers to deployment. In Proceedings of the European Battery, Hybrid
and Fuel Cell Electric Vehicle Congress, pages 1–8, Geneva, Switzerland, 2017.
[20] T.-H. Jin, H. Park, M. Chung, K.-Y. Shin, A. Foley, and L. Cipcigan. Review of virtual power
plant applications for power system management and vehicle-to-grid market development.
Trans. Korean Inst. Electr. Eng., 65(12):2251–2261, 2016. doi:10.5370/KIEE.2016.
65.12.2251.
[21] T. Lehtola and A. Zahedi. Electric vehicle to grid for power regulation: A review. In 2016
IEEE International Conference on Power System, Wollongong, NSW, Australia, 2016. IEEE.
doi:10.1109/POWERCON.2016.7753880.
[22] Repsol. Quantos kwh consome uma família. https://www.repsol.pt/

particulares/assessoramento/quantos-kwh-consome-uma-familia/.
[23] S. A. Gorji, M. Ektesabi, and J. Zheng. Isolated switched-boost push–pull DC–DC converter
for step-up applications. IET Electronics Letters, 53(3):177–179, 2017. doi:10.1049/
el.2016.4151.
[24] A. A. Aboulnaga and A. Emadi. Performance evaluation of the isolated bidirectional Cuk
converter with integrated magnetics. In Proc. IEEE APEC, volume 2, pages 1557–1562,
Aachen, Germany, Feb. 2004. doi:10.1109/PESC.2004.1355657.
[25] Gustavo G. Koch, Samuel S. Queiroz, Cassiano Rech, Ricardo C. L. F. Oliveira, Renato A.
Borges, Eduardo S. Tognetti, and Vinícius F. Montagner. Design of a robust pi controller for
a dual active bridge converter. In 2016 12th IEEE International Conference on Industry Ap-
plications (INDUSCON), pages 1–6, 2016. doi:10.1109/INDUSCON.2016.7874503.
REFERENCES 63
[26] S. Talbi, A. M. Mabwe, and A. E. Hajjaji. Control of a bidirectional dual active bridge
converter for charge and discharge of a Li-ion battery. In Proc. IEEE IECON, pages 849–
856, Yokohama, Japan, Nov. 2015. doi:10.1109/IECON.2015.7392205.
[27] F. Krismer and J. W. Kolar. Accurate small-signal model for the digital control of an automo-
tive bidirectional dual active bridge. IEEE Transactions on Power Electronics, 24(12):2756–
2768, Dec. 2009. doi:10.1109/TPEL.2009.2027904.
[28] S. Shao, L. Chen, Z. Shan, F. Gao, H. Chen, D. Sha, and T. Dragicevic. Modeling and
advanced control of dual-active-bridge dc–dc converters: A review. IEEE Access, 7:108280–
108305, 2019. doi:10.1109/TPEL.2021.3108157.
[29] S. Q. Zhao, W. Liu, and Y. Sun. Overview of dual-active-bridge isolated bidirectional DC-DC
converter for high-frequency-link power-conversion system. IEEE Transactions on Power
Electronics, Aug. 2014. doi:10.1109/TPEL.2013.2289913.
[30] Haihua Zhou, Tran Duong, Siew Tuck Sing, and Ashwin M. Khambadkone. Interleaved
bi-directional dual active bridge dc-dc converter for interfacing ultracapacitor in micro-grid
application. In 2010 IEEE International Symposium on Industrial Electronics, pages 2229–
2234, 2010. doi:10.1109/ISIE.2010.5636841.
[31] W. Liu, T. Placke, and K.T. Chau. Overview of batteries and battery management for electric
vehicles. Energy Reports, 8:4058–4084, 2022. doi:10.1016/j.egyr.2022.03.016.
[32] Infineon Technologies.

FF300R12KS4 datasheet. https://pdf1.alldatasheet.
com/datasheet-pdf/view/416949/INFINEON/FF300R12KS4.html.
[33] Infineon Technologies. FF400R06KE3 datasheet. https://www.infineon.com/cms/

en/product/power/igbt/igbt-modules/ff400r06ke3/.
[34] Philips. CD4050 hex buffer non-inverter IC datasheet. https://components101.com/

ics/cd4050-hex-buffer-non-inverter-ic-datasheet-pinout-features.
[35] Semikron. Sk25gh063 datasheet. https://datasheetspdf.com/pdf-file/

713580/Semikron/SK25GH063/1.
[36] BK Precision. Bk precision 8500 series programmable DC electronic loads. https://

www.bkprecision.com/products/vids/8500.
[37] Zhe Zhang and Michael Andersen. High frequency ac inductor analysis and design for dual
active bridge (dab) converters. 03 2016. doi:10.1109/APEC.2016.7468006.
[38] Shuvankar Dey, Surja Sekhar Chakraborty, Saurabh Singh, and Kamalesh Hatua. Design of
high frequency transformer for a dual active bridge (dab) converter. In 2022 IEEE Global
Conference on Computing, Power and Communication Technologies (GlobConPT), pages
1–6, 2022. doi:10.1109/GlobConPT57482.2022.9938249.

DAB_1714901843

Uploaded by

Copyright:

Available Formats

DAB_1714901843

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

DAB_1714901843

Uploaded by

Copyright:

Available Formats

FACULDADE DE E NGENHARIA DA U NIVERSIDADE DO P ORTO

Dual Active Bridge Converter for

André Filipe Gama de Sousa Rodrigues

Mestrado em Engenharia Eletrotécnica e de Computadores

Supervisor: António José de Pina Martins

October 29, 2023

2 State of the Art 3

3 Background on the Dual Active Bridge 17

6 Conclusions and Future Work 59

3.1 DAB converter circuit [6] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4.1 DAB circuit on PSIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.10 Waveforms for 30 degree phase shift. . . . . . . . . . . . . . . . . . . . . . . . . 33

5.1 STM32F429 Development Board. . . . . . . . . . . . . . . . . . . . . . . . . . 42

2.1 Differences between AC and DC charging. . . . . . . . . . . . . . . . . . . . . . 7

3.1 Operating modes of the DAB . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4.1 Simulation value table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

5.1 Internal Trigger Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

1.1 Context and Motivation

• Chapter 4 details the PSIM simulation’s implementation and results.

State of the Art

2.1 World Electricity Production

2.2 Electric Vehicle Shift and EV Charging Stations

Figure 2.2: Global greenhouse gas emissions by sector [2]

Figure 2.3: AC vs DC charging diagram. [3]

2.2.2 Levels of EV charging

2.3 Vehicle-to-Grid Technology

2.4 Topology of Bidirectional DC-DC Converters

Figure 2.6: Diagram of bidirectional EV charging stations in grid [5]

Figure 2.7: DC-DC topology flowchart [5]

2.4.1 Isolated Converters

2.4.1.1 Bidirectional Flyback Converter

Figure 2.8: Bidirectional Flyback DC-DC Converter [5]

2.4.1.2 Isolated Cuk & Sepic/Zeta Bidirectional Converter

2.4.1.3 Push-Pull Bidirectional Converter

Figure 2.10: Push-Pull Bidirectional DC-DC Converter [5]

2.4.1.4 Forward Bidirectional Converter

Figure 2.11: Forward Bidirectional DC-DC Converter [5]

2.4.1.5 Dual Half-Bridge Bidirectional Converter

Figure 2.12: Dual Half-Bridge Bidirectional DC-DC Converter [5]

2.4.1.6 Half Bridge-Full Bridge Bidirectional Converter

Figure 2.13: Half Bridge-Full Bridge Bidirectional DC-DC Converter [5]

2.4.2 Control Schemes for Bidirectional Converters

Figure 2.14: PID control feedback diagram

Figure 2.15: Sliding mode control feedback diagram

Figure 2.16: Fuzzy control feedback diagram

Figure 2.17: Digital control feedback diagram

Control Control Benefits Limitations

Background on the Dual Active Bridge

3.1 Theoretical Development

Figure 3.1: DAB converter circuit [6]

Table 3.1: Operating modes of the DAB

Operating Modes Transistor Pairs Conducting

Figure 3.3: DAB Current flow at each operating mode.

Figure 3.4: PI controller diagram for DAB applications

3.4 Interleaved Dual Active Bridge

4.2 Specifications and Circuit Setup

Table 4.1: Simulation value table.

Figure 4.1: DAB circuit on PSIM.

Figure 4.2: Interleaved DAB circuit on PSIM.

4.3.1 Phase Shift Control Implementation