Electronics 09 00887 v2 PDF

electronics
Review
Brushed DC Motor Drives for Industrial and
Automobile Applications with Emphasis on
Control Techniques: A Comprehensive Review
Dimitris A. Barkas *, George C. Ioannidis, Constantinos S. Psomopoulos , Stavros D. Kaminaris
and Georgios A. Vokas
Department of Electrical & Electronics Engineering, University of West Attica, 122 44 Egaleo, Greece;
gioan@uniwa.gr (G.C.I.); cpsomop@uniwa.gr (C.S.P.); skamin@uniwa.gr (S.D.K.); gvokas@uniwa.gr (G.A.V.)
* Correspondence: d.barkas@uniwa.gr; Tel.: +30-210-538-1541

Received: 21 April 2020; Accepted: 25 May 2020; Published: 27 May 2020
Abstract: The current paper presents an inclusive survey about the AC to DC and DC to DC
converters for brushed DC Motor Drives. An essential number of different AC to DC and DC to
DC topologies and control techniques, applied on the brushed DC motor drives are presented. This
extensive literature review exposes advantages, disadvantages and limitations besides giving the
basic operating principles of various topologies and control techniques.
Keywords: dc motor drives; hard switching; soft switching; controlled-rectifiers; choppers; fuzzy
logic controllers; artificial neural network controllers
1. Introduction
DC motors have been used for several decades. The DC network was the first developed electric
network and was constructed to work on the DC electric network. Nowadays, the majority of industry
installed motors consist of AC motors, due to their high-speed operation and their smaller volume
and weight. Additionally, AC motors, due to their construction, require lighter maintenance and are
cheaper compared to the DC motors. However, DC motors are still used for several reasons, including,
wide speed range, starting and accelerating torques more than 400% of their rated values, good speed
regulation and simpler and cheaper control systems. Their main applications, include manufacture of
pulp, paper and paperboard, propulsion of electric vehicles, textile industries and public transportation,
such as subway and trolley systems. Modern DC motor drives utilize power electronic devices and are
subdivided to chopper-fed and controlled thyristor-fed drives.
DC motor drives can be categorized according to the way they manage the energy generated
during braking of the DC motor [1–4]. From that point of view, there are non-regenerative and
regenerative DC drives in industry. Non-regenerative DC drives are the most conventional type widely
used. They have an ability to control motor speed and torque only in one direction, which means
operation in the first-quadrant of the four quadrants of operations, as illustrated in Figure 1.
The controller output polarity can be reversed using a magnetic armature reversing contractor
or using a manual switch. By operating the controller in this way, the motor armature direction is
also reversed. The DC motor operates in the third quadrant of Figure 1. In both cases, torque and
rotational direction are the same. Four-quadrant drives, or as usually referred regenerative DC drives,
achieve to control both speed and direction of motor rotation. Moreover, regenerative DC drives can
also control the direction of motor torque. These DC drives are named as “regenerative” due to the
regeneration ability that they offer. The regeneration feature is the ability of the drive to convert the
mechanical energy produced during the braking condition, into electrical energy, which is returned to
Electronics 2020, 9, 887; doi:10.3390/electronics9060887 www.mdpi.com/journal/electronics

Electronics 2020, 9, 887 2 of 26
Electronics 2019, 8, x FOR PEER REVIEW 2 of 27
the power source.

is returned Duringsource.
to the power the drive operation
During in the
the drive first andin
operation third quadrants,
the first thequadrants,
and third motor rotation and
the motor
motor torque have got the same direction and the drive operates as a conventional drive without
rotation and motor torque have got the same direction and the drive operates as a conventional drive the
regeneration
without the capability.
regeneration capability.
Figure1.1.Quadrants
Figure Quadrantsofofoperation
operationofofaaDC
DCmotor.
motor.
The
Thespecial
special capabilities
capabilities of regenerative drives
of regenerative drivesexistexistonlyonlyduring
duringthe the operation
operation in the
in the second
second and
and fourth quadrants. Operating in these quadrants, a braking force
fourth quadrants. Operating in these quadrants, a braking force is provided due to the motor torque, is provided due to the motor
torque,
which which
opposes opposes
to theto direction
the direction of of motor
motor rotation.One
rotation. Onecharacteristic
characteristic of of aahigh-performance
high-performance
regenerative
regenerative drive is its ability to rapidly switch from braking to motoring modesand
drive is its ability to rapidly switch from braking to motoring modes andvice
viceversa,
versa,
controlling
controlling simultaneously the direction of motor rotation. A DC regenerative drive is practicallyaa
simultaneously the direction of motor rotation. A DC regenerative drive is practically
combination
combinationofof twotwoDCDC drives, which
drives, offer the
which offeroperation of the motor
the operation of thein motor
the firstinand thefourth
first quadrants,
and fourth
orquadrants,
in the second and third quadrants, respectively.
or in the second and third quadrants, respectively.
The
Theclassification
classificationofofDC DCmotor
motordrives
drivescancanbe bedone
doneaccording
accordingtotothe thetype
typeof ofthe
theutilized
utilizedconverter,
converter,
which
whichcontrols
controlsthethe speed
speed and and the
thetorque
torqueofofthe theDCDC motor
motor [2–4].
[2–4]. When When a controlled
a controlled one-phase
one-phase or
or three-
three-phase rectifier is used, then this drive category is called Controlled Rectifier-Fed
phase rectifier is used, then this drive category is called Controlled Rectifier-Fed or Thyristor-Fed DC or Thyristor-Fed
DC motor
motor drive.
drive. Another
Another DC DC motor
motor drivecategory
drive categoryisisthe theChopper-Fed,
Chopper-Fed, in in which
which aa DC DC totoDCDCconverter
converter
isisused.
used.TheTheprevious
previous categories
categories cancan be be
further subdivided
further subdivided intointo
non-regenerative
non-regenerative and regenerative
and regenerative DC
motor drives according to the way they manage energy during
DC motor drives according to the way they manage energy during braking condition. braking condition.
There
Thereareareaalarge
largenumber
numberofofconverters
convertersthat thathave
havebeen beenpresented
presentedand andanalyzed
analyzedin inthe
theliterature.
literature.
These converters operate using controlled rectifier circuits and they can
These converters operate using controlled rectifier circuits and they can be categorized according to be categorized according
tothe
theinput
inputpower,
power,asassingle-phase
single-phaseororthree-phase,
three-phase, AC AC totoDC DC converters
converters [2–9].
[2–9]. The The application
application of of
the
the above-mentioned
above-mentioned converters
converters variesaccording
varies accordingtotothe the special
special requirements.
requirements. These These converters
converterscan can
operate
operateatatone,one,two
twoororthree threequadrants.
quadrants.However,
However,the thefabrication
fabricationofofmore moreefficient
efficientsemiconductor
semiconductor
switches,
switches,during
duringlastlastdecades,
decades, has has ledled to
to the development of new new DC DC to to DCDCconverters.
converters.These TheseDC DCto
toDCDC converters
converters have
have thethe ability
ability to operate
to operate at high
at high frequencies,
frequencies, particularly
particularly in thein the range
range of tensofoftens
kHz.
ofThey
kHz.alsoThey also high
exhibit exhibit high response
response speed and speed
theirand their application
application for controlling
for controlling DC motors DCismotors
increased is
increased
[2–5,10–26].[2–5,10–26].
During theDuring recent theyears,recent
thereyears,
has alsothere beenhasa also
trend been a trend in
in reducing thereducing the size,
size, volume, and
volume, and weight of the drive. Additionally, the requirement of controlling
weight of the drive. Additionally, the requirement of controlling DC motors with low inductance, led DC motors with low
inductance, led to theofdevelopment
to the development converters which of converters
operate which
with even operate withfrequency
higher even higher andfrequency
reduced rippleand
reduced
current.ripple current.
For these reasons,For DCtheseto reasons, DC to were
DC converters DC converters
used [11–26]. were Atused
first [11–26].
these DCAt tofirst these DC
DC converters
towere
DC applied
converters were applied
in switching power in supplies
switching powerand
[11–15] supplies
they could[11–15]
not and they could
be applied for DC notmotor
be applied
drives,
for DC motor
especially fordrives,
electricespecially for electric They
vehicle propulsion. vehicle propulsion.
suffer They suffer
from excessive voltagefrom excessive
and currentvoltage
stresses
and current
[13,15], andstresses
moreover [13,15],
they and moreover
are unable they are
to handle theunable
backward to handle
powerthe whichbackward powerduring
is generated which the is
generated during the regenerative braking [16]. In the literature, another
regenerative braking [16]. In the literature, another category of drives has been proposed which are category of drives has been
proposed which
suitable for DC are
motorsuitable
drives, forand
DCthismotor drives,isand
category known this category is knownDC
as soft-switching as soft-switching
to DC converters DC[17–to
DC converters
26]. [17–26].
Additionally,
Additionally,the the speed
speed control of ofaaDC DCmotor
motoris is achieved
achieved using
using armature
armature control
control methods,
methods, field
field control
control methods
methods and and combination
combination of them.
of them. In armature
In armature control
control method method the field
the field current
current is keptis
kept approximately constant and the armature current
approximately constant and the armature current is varied by a control is varied by a control signal provided by
provided by the the
controller, a PI controller for example. In this method, the field current is kept constant which means
that flux density is constant resulting to a constant field flux in the field circuit. Therefore, the motor
Figure 2 [27].
Considering the field control methods, the most popular ones are the field voltage control with
constant armature current and the field voltage control with constant armature voltage [28]. In the
field voltage control with constant armature current, the use of a current generator is mandatory, and
this is a 2020,
Electronics basic disadvantage due to the high cost of such a generator, especially for high power
9, 887 3 of 26
applications. Additionally, as the rotor current maintains constant, a dangerous situation for the
motor’s life resulted, due to the high values of current in rotors that may occur. However, as the
controller,
power input a PI controller
maintains for example.
constant In this
and if the method,
control device the field acurrent
drives is kept
low power constant
device, thenwhich means
a low power
that flux density is constant resulting to a constant field flux in the field circuit.
reference signal is required to control the system instead of the use of a high precision with high Therefore, the motor
gain
torque directly
amplifier. dependsusing
Additionally, on armature current
this control which
method, can motors
many be easily varied
could by varyinginthe
be connected armature
series when
avoltage.
high-power In contrast,
currentingenerator
combinedis armature
available,and fieldspecific
under controlsystem
(obviously for separately
configuration excited
[28]. In DC
the field
motor), the speed can be varied from zero to rated value mainly by armature voltage
voltage control with constant armature voltage, the constant voltage generator is usually cheap, even control in the
constant
in the case torque and linearly
of high-power varyingAdditionally,
purposes. power region. whenThetheincrease
powerofisthe DC motor
supplied speed
to the systemabove
usingits
rated value can be achieved by reducing field flux in the so-called constant power
the constant voltage generator then the control signal is a low power signal. Moreover, as the usual and hyperbolic
torqueare
cases region. The combined
balanced armature armature and field
voltage using speed control
a non-null back with
EMF torque and power are
(Electro-Magnetic shown
Force), in
this
Figure 2 [27].
configuration results to a limited number of peaks on the armature current [28].
Figure 2. Combined
Combined armature
armature and field control of DC motor [27].
Considering
The the field of
implementation control methods,
a complete DCthe mostdrive
motor popular ones are
involves the the field voltage
development of control with
a controller
constant
which armature
will drive thecurrent
powerand the field
circuit. There voltage control
are many with constant
available techniques armature voltagesuch
to implement [28]. action,
In the
field the
with voltage
mostcontrol
popularwith constant armature
categorization to be: current, the use of a current generator is mandatory,
and this is a basic disadvantage due to the high cost of such a generator, especially for high power
 Classical PID Controllers
applications. Additionally, as the rotor current maintains constant, a dangerous situation for the
 Intelligent controllers.
motor’s life resulted, due to the high values of current in rotors that may occur. However, as the
powerThe work
input presented
maintains by the and
constant authors
if theincontrol
this manuscript is ana effort
device drives of extensive
low power device, literature
then a lowreview,
power
exposing
reference advantages, disadvantages
signal is required to control and limitations,
the system besides
instead of thegiving
use ofthe basicprecision
a high operating principles
with high gainof
various topologies and control techniques. Given the fact that this work is actually
amplifier. Additionally, using this control method, many motors could be connected in series when a review paper,
there is no set method
a high-power current to ensure that
generator all the literature
is available, under available globally
specific system on the topic, [28].
configuration was taken
In theunder
field
consideration
voltage controland theconstant
with chancesarmature
of the review
voltage,being
thebiased.
constantThe authors,
voltage firstly, present
generator is usuallythecheap,
topologies
even
of DC case
in the motor drives, and purposes.
of high-power secondly, Additionally,
they review the whencontrol techniques
the power basedtoon
is supplied thethe mostusing
system popularthe
categorization
constant voltage already mentioned
generator then theabove.
control signal is a low power signal. Moreover, as the usual cases
are balanced armature voltage using a non-null back EMF (Electro-Magnetic Force), this configuration
2. DC Motor
results Drivesnumber
to a limited Topologies
of peaks on the armature current [28].
The implementation of a complete DC motor drive involves the development of a controller which
will drive the power circuit. There are many available techniques to implement such action, with the
most popular categorization to be:
â Classical PID Controllers

â Intelligent controllers.
The work presented by the authors in this manuscript is an effort of extensive literature review,
exposing advantages, disadvantages and limitations, besides giving the basic operating principles of
various topologies and control techniques. Given the fact that this work is actually a review paper,
there is no set method to ensure that all the literature available globally on the topic, was taken under
consideration and the chances of the review being biased. The authors, firstly, present the topologies
of DC motor drives, and secondly, they review the control techniques based on the most popular
categorization already mentioned above.
2. DC Motor Drives Topologies
As
As it
it was
was mentioned
mentioned before,
before, DC
DC motor
motor drives
drives are
are divided
divided into
into Controlled
Controlled Rectifier-Fed
Rectifier-Fed and
and
Chopper-Fed
Chopper-Fed DC
DC Motor
Motor Drives.
Drives.
2.1. Controlled
Controlled Rectifier-Fed
Rectifier-Fed DC Drives
One of the main applications
applications ofof the
the thyristor-based
thyristor-based DC drive is the the control
control of DC motors,
motors, which
are equipped with brushes, in applications where the maintenance is higher but tolerable. The speed
control isisachieved
achievedbased
basedononthethe
feedfeed
of a of a low-impedance
low-impedance adjustable
adjustable DC voltage
DC voltage in the
in the motor motor
armature
armature
through the through the thyristor-based
thyristor-based rectifier. Although
rectifier. Although the usualthe usual
case is a case is a three-phase
three-phase system tosystem to
feed the
feed the rectifier with power, when the motor power is up to a few kilowatts, a single-phase
rectifier with power, when the motor power is up to a few kilowatts, a single-phase system can be system
can
used.beOn
used.the On thehand,
other otherinhand,
orderintoorder
supplytothe
supply
field the field winding
winding witha energy,
with energy, a single-phase
single-phase system is
system
used asistheused as the
needed needed
power powerlower
is much is much lower comparing
comparing to the power
to the armature armature power
(Figure 3).(Figure 3).
Figure 3.
Figure General closed-loop
3. General closed-loop controlled
controlled rectifier-fed
rectifier-fed DC motor drive.
The main
The mainpower
powercircuit
circuitusually
usuallyconsists
consistsofofoneoneor or four
four or six-thyristor
or six-thyristor circuit,
circuit, whichwhich rectifies
rectifies the
the incoming AC supply to produce a DC supply to the motor armature.
incoming AC supply to produce a DC supply to the motor armature. Altering the firing angle of Altering the firing angle
the
of the thyristor/s
thyristor/s the mean
the mean valuevalue
of theof rectified
the rectified voltage
voltage cancan be be varied
varied andandthus,
thus,thethemotor
motor speed
speed isis
controlled. The controlled rectifier produces a DC bus with an explicit ripple
controlled. The controlled rectifier produces a DC bus with an explicit ripple in the output voltage. in the output voltage.
This ripple
This ripple component
component provokes
provokes pulsating
pulsating current
current andand fluxes
fluxes inin the
the motor,
motor, andand in in order
order toto avoid
avoid
excessive eddy-current
excessive eddy-currentlosseslossesandandcommutation
commutation problems,
problems, thethe
motor poles
motor and frame
poles and frameshould be made
should be
of laminated material. It is a common practice that motors, supplied for
made of laminated material. It is a common practice that motors, supplied for use with thyristor use with thyristor drives,
should should
drives, have laminated construction.
have laminated Nevertheless,
construction. older motors
Nevertheless, often have
older motors oftensolid
havepoles
solidand/or frames
poles and/or
so these motors will not always work satisfactorily with a rectifier-based
frames so these motors will not always work satisfactorily with a rectifier-based power supply. power supply. It is also
It is
common that drive motors are supplied with an attached blower motor
also common that drive motors are supplied with an attached blower motor as standard. In this way, as standard. In this way,
continuous ventilation
continuous ventilation is
is provided
provided andand the
the motor
motor can can operate
operate continuously
continuously at at full
full torque
torque andand lowest
lowest
speeds without overheating
speeds without overheating [2–6]. [2–6].
Motor current
Motor current and
and speed
speed are
are the
the basic
basic variables
variables of of interest
interest which
which are are monitored
monitored usingusing low
low power
power
control circuits.
control circuits. These
Thesevariables
variablesare areused
usedin in
order to generate
order to generate appropriate
appropriate firingfiring
pulses so thatsothe
pulses motor
that the
maintains constant speed despite load variation. The speed reference
motor maintains constant speed despite load variation. The speed reference signal, as shown in signal, as shown in Figure 3,
is typically
Figure 3, is an analogue
typically an voltage
analogue varying
voltagefrom 0 to 10from
varying V or 0from
to 100 to or −50to
V 5orVfrom to55V,Vand obtained
or −5 to 5 V,from
and
an appropriate speed-setting potentiometer or from elsewhere in the
obtained from an appropriate speed-setting potentiometer or from elsewhere in the system. The system. The power topology,
the control
power and protective
topology, circuits
the control andconstitute
protective thecircuits
converter. Standard
constitute themodular
converter.converters
Standardare modular
available
as off-the-shelf
converters items from
are available as 0.5 kW up toitems
off-the-shelf several hundred
from 0.5 kWkW, up while larger
to several drives will
hundred kW,be tailored
while to
larger
specific requirements. Individual converters may be mounted in enclosures
drives will be tailored to specific requirements. Individual converters may be mounted in enclosures with isolators, fuses etc.,
or groups
with of converters
isolators, fuses etc.,may be mounted
or groups together to
of converters may formbe amounted
multi-motor drive.to form a multi-motor
together
drive.
Figure 4 presents a separately excited DC motor. In this case, the supply is fed using a single-
phase half wave converter [2–6]. This type of drives can drive the DC motor permitting its usage only
in one quadrant, and it is used for small DC motors with rating power up to 0.5 kW. The average
output voltage for this converter can be found to be,
Vm
V = (1 + cosα), for 0 < α < π (1)
where V is the maximum value of the applied line voltage, and a is the firing angle.
Figure 5 illustrates a separately excited motor which is supplied with electric power through a
single-phase asymmetrical converter. As the diodes in the structure of Figure 5 cannot have at any
instant a positive voltage across their terminals, it means that the motor’s armature voltage cannot be
Electronics
negative.2020, 9, 887
Consequently, this topology cannot be regenerative and it can operate only to the5 of 26
first
quadrant of the torque-speed diagram. Additionally, its usage is limited for DC drives with total
power up to 15 kW. In order to protect the circuit from overvoltages, the used diodes should be an
Figure 4 presents a separately excited DC motor. In this case, the supply is fed using a single-phase
ultra-high-speed type, aiming to offer a trusted freewheeling path for the main power circuit. In the
half wave converter [2–6]. This type of drives can drive the DC motor permitting its usage only in
case of one-quadrant operation the fully controlled converters exhibit worst characteristics, in
one quadrant, and it is used for small DC motors with rating power up to 0.5 kW. The average output
comparison to the half-controlled. Specifically, they have more harmonic distortion to the input
voltage for this converter can be found to be,
current, more expensive control circuits, lower power factor and decreased mean value of the output
voltage for the same firing angle. If V this
m converter type is in continuous operation mode, then the
Vo = (1 + cos α), for 0 < α < π (1)
average output voltage can be found 2π to be [2–6]:
Vm
where Vm is the maximum value of V the (1 +line
o =applied cosα), for 0and
voltage, < α a< π.
is the firing angle. (2)
Electronics 2019, 8, x FOR PEER REVIEW π 5 of 27
where V is the maximum value is = iT of the applied line voltage, iα and a is the firing angle.
T
Figure 5 illustrates a separately excited motor which is supplied Rα with electric power through a
single-phase asymmetrical converter. As the diodes in the structure of Figure 5 cannot have at any
instant a positive voltage across their terminals, it means that the motor’s armature voltage cannot be
negative. Consequently, thisν topology cannot be Vo La and it can operate only to the first
s = ν m sin ω t
FDregenerative
quadrant of the torque-speed diagram. Additionally, its usage is limited for DC drives with total
power up to 15 kW. In order to protect the circuit from overvoltages, the used diodes should be an
ultra-high-speed type, aiming to offer a trusted ifreewheeling
fd eb path rfor
f , L fthe V
main
f power circuit. In the
case of one-quadrant operation the fully controlled converters exhibit worst characteristics, in
comparison to the half-controlled. Specifically, they have more harmonic distortion to the input
current, more expensive control circuits, lower power factor and decreased mean value of the output
Figure
Figure 4. 4. Single-phase
Single-phase half
half wave
wave converter
converter drive.
drive.
voltage for the same firing angle. If this converter type is in continuous operation mode, then the
average output
Figure voltage can
5 illustrates be found excited
a separately to be [2–6]:
motor which is supplied with electric power through a
Vi m T1 iα
single-phase asymmetrical converter. Vo = TAs (1
the
11
diodes
+ cosα),in for
D the0structure
11
<Rαα < π. of Figure 5 cannot have at any (2)
instant a positive voltage across their terminals, π it means that the motor’s armature voltage cannot
be negative. Consequently, this topology cannot be regenerative L
and it can operate only to the first
a
is = idiagram. V
quadrant of the torque-speed T T Additionally, iits
α usage
o
is limited for DC drives with total
power up to 15 kW. In order νto= νprotectsin ωt
s m
Rα
the circuit from overvoltages, the used diodes should be an
e r ,L
b V f f f
ultra-high-speed type, aiming to offer a trusted T
freewheeling
D 12
path for the main power circuit. In the case
12
of one-quadrant operation the fully controlled converters exhibit worst characteristics, in comparison
Vo La
ν s = ν m sinthey
to the half-controlled. Specifically, ωt have more FD harmonic distortion to the input current, more
Figure lower
expensive control circuits, 5. Single-phase
power factorhalf-controlled asymmetrical
and decreased mean value converter drive.
of the output voltage for the
same firing angle. If this converter type is in continuous operation mode, then the average output
In Figure 6, a full wave converter is illustrated. eb
i fd In this converter rf , Ltype,
f Vthe
f
armature voltage upper
voltage can be found to be [2–6]:
limit is +2 Vm/π and the lower voltage limit is −2 Vm/π. Consequently, this converter type operates in
the first and fourth quadrants, and its Vmusage is limited for motors up to 15 kW. The output voltage
Vo = (1 + cos α), for 0 < α < π. (2)
for the full wave converter is described π using the following equation [2–6]:
Figure 4. Single-phase2Vm half wave converter drive.
Vo = cosα, for 0 < α < π. (3)
π
i iα
Figure 7 depicts a single-phase fully-controlled dualT1converter drive. In this version, there is a
T D 11 Rα 11
back-to-back connection of two single-phase full wave converters. The first full wave converter
supplies the armature with a positive voltage +Vo, and has the ability to operate in the first and fourth
L
quadrants. The second full wave converter feeds the V armature with a negative voltage −Vo and
o
a
operates in the second and νthird quadrants. Obviously, this single-phase dual converter is a four-
= ν sin ωt
s m
quadrant converter providing the four modes of operations, e rwhich
,L V are the motoring mode, the
b f f f
T D 12
forward braking mode or forward regeneration mode, the reverse motoring mode and the reverse
12
braking mode or reverse regeneration mode.

Figure
Figure 5.
5. Single-phase
Single-phasehalf-controlled
half-controlled asymmetrical
asymmetrical converter
converter drive.
drive.
In Figure
In Figure 6,
6, aafull
fullwave
waveconverter
converter is
isillustrated.
illustrated. In
Inthis
thisconverter
convertertype,
type,the
thearmature
armaturevoltage
voltage upper
upper
limit is +2 V /π and the lower voltage limit is −2 V /π. Consequently, this converter type operates
limit is +2 Vmm/π and the lower voltage limit is −2 Vmm/π. Consequently, this converter type operates in in
the first and fourth quadrants, and its usage is limited for motors up to 15 kW. The output voltage
for the full wave converter is described using the following equation [2–6]:
2Vm
Vo = cosα, for 0 < α < π. (3)
π
Figure 7 depicts a single-phase fully-controlled dual converter drive. In this version, there is a
the first and fourth quadrants, and its usage is limited for motors up to 15 kW. The output voltage for
the full wave converter is described using the following equation [2–6]:
2Vm
Vo = cos α, for 0 < α < π. (3)
π
iT 11 iα
T11 T13 Rα
Vo La
ν s = ν m sin ωt
eb rf , L f Vf
T14 T12
iT 11 fully-controlled
Figure 6. Single-phase iα converter drive.
T11 T13 Rα
Figure 7 depicts a single-phase fully-controlled iα
dual converter drive. In this version, there is
iT 11
a back-to-back connection of two single-phase fullRwave converters. La T
The first full wave converter
T11 T13 Vαo T24
supplies the armature with a positive voltage +Vo , and has the ability to operate in the first and
22
Vo
fourth quadrants. The second ν s = νfull wave converter feeds the armature with a negative voltage −Vo
m sin ω t
La eb rf , L f
and operates in the second and third quadrants. Obviously, this Vf
single-phase dual converter is a
T14 T12
four-quadrant converter providing the four modes of operations, which are
ν s = ν m sin ωt the motoring mode, the
ν s = ν m sin ωt
forward braking mode or forward regeneration mode, eb the reverse motoring mode and the reverse
T14 T12 rf , L f T23 T21
braking mode or reverse regeneration
Figure 6. Single-phase mode. fully-controlled converter drive.
Vf
T 11 i iα
Figure 7. Single-phase fully-controlled dual converter drive.
T11 T13 Rα T22 T24
Vo
The output voltage for the first, assuming that it operates at a firing angle α1, is [2–6]:
Vm La
Vo = (1 + cosα1 ), for 0 < α < π. (4)
π
ν s = ν m sin ωt ν s = ν m sin ωt
In the same way, the output voltage for the second
eb converter which operates at a firing angle α2
T14 T12 rf , L f T23 T21
then is [2–6]:
Vf
Vm
Vo = (1 + cosα2 ). (5)
π
The inverse operation, in this case, could be achieved through a field-reversal, which means that
The
the motor’s output voltage
counter for the first, assuming
electromotive that it operates at a firing anglethis
α1 , is [2–6]: is finally
The output voltage for the first,force (CEMF)
assuming will
that be reversed.
it operates However,
at a firing angle α1, inversion
is [2–6]:
achieved with the inversion in the motor’s V armature current. The inversion through the armature’s
VoV=m rapid
m
current presents the advantageVof o =more (1 +(1cos
+ cosα
α 1 )1, ),for
reversal. for
In 0<< α α <<π.the
0addition, π. DC motor drives provide the (4)
π π
fastest dynamic response to changes in torque or speed commands. One desirable characteristic,
In the same way, the output voltage for the second converter which operates at a firing angle α2
whichIn leads
the sameto a way, the output
satisfactory voltageoffor
operation thethe second
control converter
system is thewhich operates
armature at acontinuity.
current firing angleTheα2
then is [2–6]:
then is [2–6]:
continuity of armature current is a desirable feature for the satisfactory operation of the control
VVmm (1 + cosα
system. Where an additional inductance VoV=o is= included
(1 + cosin α22)a).. series with the motor armature, the (5)
continuous-current operation can be achieved, ππwithout the certainty that a quite large inductor will
lead The
to a inverse this case,
operation, inoperation
continuous-current this case, for
could
could be achieved
all be
loadachieved
conditions through
through aa field-reversal,
field-reversal, which means that
and speed.
the motor’s
the motor’s counter
counter electromotive
electromotive force
force (CEMF)
(CEMF) will
will
The size of the necessary induction could be reduced by employing be
be reversed.
reversed. However,
However, this inversion
this inversion is finally
is finally
other converter
achieved
achieved with the
with
configurations, inversion
such in the motor’s
as three-phase AC toarmature The inversion
current.However,
DC converters. The inversion through
thesethrough the armature’s
configurations do not
current presents the advantage of more rapid
ensure the complete elimination of the inductor’s size. reversal. In
In addition,
addition, the
the DC motor drives provide the
fastest
Adynamic
dynamic
three-phaseresponse
response totochanges
changes
fully-controlled in torque
in torque
converter orisspeed
or commands.
speed
illustrated commands.
in FigureOne8desirable
One characteristic,
[2–6].desirable
This type which
characteristic,
of converter
leads
workstoleads
which ina satisfactory
twoto quadrantsoperation
a satisfactory of the control
operation
of operation of the
and system
controlissystem
traditionally theisarmature
is the
used current
forarmature continuity.
large motorcurrent The continuity
continuity.
applications Thea
with
continuity of armature current is a desirable feature for the satisfactory
load up to 140 kW. A variation of this converter type can be achieved, which replaces the three operation of the control
system.
thyristors Where an additional
by diodes. inductance
This configuration is included
is limited to theinfirst a series
quadrant withof the motor and
operation armature, the
introduces
continuous-current operation can be achieved, without the certainty that
even harmonics into the line currents making prohibitive its application on large power applications. a quite large inductor will
lead
The to a continuous-current
output operation for
voltage of the three-phase all load
fully conditions
controlled converterand speed.
in continuous-mode operation is
givenThe sizefollowing
by the of the equation
necessary[2–6], induction could be reduced by employing other converter
of armature current is a desirable feature for the satisfactory operation of the control system. Where an
additional inductance is included in a series with the motor armature, the continuous-current operation
can be achieved, without the certainty that a quite large inductor will lead to a continuous-current
operation for all load conditions and speed.
The size of the necessary induction could be reduced by employing other converter configurations,
such as three-phase AC to DC converters. However, these configurations do not ensure the complete
elimination of the inductor’s size.
A three-phase fully-controlled converter is illustrated in Figure 8 [2–6]. This type of converter
works in two quadrants of operation and traditionally is used for large motor applications with a load
up to 140 kW. A variation of this converter type can be achieved, which replaces the three thyristors by
diodes. This configuration is limited to the first quadrant of operation and introduces even harmonics
into the line currents making prohibitive its application on large power applications. The output
voltage of the three-phase fully controlled converter in continuous-mode operation is given by the
following equation [2–6],
Electronics 2019,
2019, 8,
8, xx FOR
FOR PEER
PEER REVIEW
REVIEW 3 Vm,L
Electronics Vo = cos α, for 0 < α < π 77 of
of 27
27
(6)
π
where V𝑽m,L
where 𝒎,𝑳 is
𝒎,𝑳 isthe
themaximum
maximumvalue
valueof
ofthe
theline
linevoltage.
voltage.
iiαα
TT1111 TT1313 TT1515 RRαα
iiaa
iibb LLaa
VVoo
iicc
eebb rrff ,, LLff Vf

V f
TT1414 TT1616 TT1212
Figure 8.
Figure 8. Three-phase
Three-phase fully-controlled converter.
Three-phase fully-controlled converter.
When four-quadrant
four-quadrant operation
operationisisneeded,
needed,then
thena adual
dualthree-phase fully-controlled
three-phase topology,
fully-controlled as
topology,
as Figure
Figure 9 illustrates,
9 illustrates, could
could be be used.
used.
iiαα
TT1111 TT1313 TT1515 RRαα TT2222 TT2626 TT2424
VVoo LLaa
eebb
TT1414 TT1616 TT1212 rrff ,, LLff TT2525 TT2323 TT2121
VV f f
Figure 9.
Figure Dual Three-phase
9. Dual Three-phase fully-controlled converter.
fully-controlled converter.
converter.
Finally,
Finally, the
the summary
summary of
of the
the key
key characteristics
characteristics of controlled rectifier-fed
of controlled rectifier-fed drives
drives are
are presented
presented in
in
Table
Table1.
1.
Table 1. Main characteristics of controlled rectifier-fed drives.
General Features
• Explicit ripple in • Low • 0.5 kW to

• 50/60 Hz operation
the output voltage order harmonics hundred kW
Single-Phase fully controlled converter
• 2Q operation
• For motors up to 15 kW
• Compared to half-controlled:
Single-phase half wave
• 1Q operation - More input current harmonic distortion
• For small DC motors up to 0.5 kW - More expensive control circuits
- Lower power factor
Single-phase asymmetrical converter
- Decreased output voltage mean value for the same
• 1Q operation firing angle
• For DC motors up to 15 kW
Single-Phase Fully-Controlled Dual converter
• 4Q operation
• Rapid reversal
• Fastest dynamic response to changes in torque or
speed commands
Three phase fully-controlled converter
Three-phase half-controlled converter
• 2Q operation
• 1Q operation • Motors up to 140 kW
• Introduces more harmonics into the
Three-phase Dual fully-controlled converter
line currents
• Its usage is prohibitive on large • 4Q operation
power applications • Rapid reversal
• Fastest dynamic response to changes in torque or
speed commands
2.2. Chopper—Fed DC Motor Drives

A chopper device is to be used when the aim is to convert a fixed DC voltage into a variable
DC output voltage. Chopper topology behaves in an identical manner as a transformer in the AC
current. Choppers are frequently used all over the world for rapid transit systems, and in some other
applications, such as mine haulers, trolley cars, forklift trucks and marine hoist. Their most significant
advantages are the regeneration operation capability, their fast response and their high efficiency. The
applied semiconductor devices in the chopper drives are the power BJT (Bipolar Junction Transistor),
the forced commutated thyristor, MOSFET (Metal-Oxid-Semiconductor Field-Effect-Transistor), IGBT
(Insulated-Gate Bipolar Transistor), MCT (MOS-Controlled Thryristor) and GTO (Gate Turn-Off
Thyristor). Independent of the applied combination, the device actually operates like a switch, which
means that when the switch is under the “on” operation, then the current flows through the load. The
on-state voltage drop on the power semiconductors is in the range of 0.5 V to 2.5 V across them, and the
power losses can be determined if their switching characteristics are also taken into consideration [2–6].
The converters can be categorized into two groups, according to the way that they transit
from one switching state to another. These converter categories are the hard-switching and the
soft-switching converters.
2.2.1. Hard-Switching Converters for DC Drives

Conventional Pulse Width Modulated (PWM) converters have been used in switched mode over
the last few decades. However, in the case of power switches, they must cut off the load current
during the turn-on and turn-off time intervals when they operate in hard-switching mode. The term
hard-switching arises due to the stressful switching situation of the power electronic devices. The
switching behavior for a hard-switching device is illustrated in Figure 10. The power device operates
during the turn-on and turn-off time intervals when they operate in hard-switching mode. The term
Electronics 2020, 9, 887
hard-switching due to the stressful switching situation of the power electronic devices.9 of
arises The26
switching behavior for a hard-switching device is illustrated in Figure 10. The power device operates
under high voltage
under high voltage and current simultaneously
and current simultaneously during
during the
the turn-or
turn-or or
or turn-off process, aa situation
turn-off process, situation
which leads into high switching stress and losses.
which leads into high switching stress and losses.
I Safe Operating Area
On Hard-switching
snubbered
Soft-switching
Off V
Figure 10. Hard and soft switching characteristics.

Figure 10. Hard and soft switching characteristics.
The reduction of stress and losses under hard-switching conditions is a phenomenon that needs to
The reduction of stress and losses under hard-switching conditions is a phenomenon that needs
be eliminated. For this reason, the addition of dissipative passive snubbers, into the power circuits, is a
to be eliminated. For this reason, the addition of dissipative passive snubbers, into the power circuits,
frequent solution, diverting by this way the stress and losses to these passive snubber circuits. Through
is a frequent solution, diverting by this way the stress and losses to these passive snubber circuits.
this addition, a reduction of the dv/dt and di/dt can be achieved. The reduction on the maximum
Through this addition, a reduction of the dv/dt and di/dt can be achieved. The reduction on the
switching frequency of the converters is also another factor aimed at reducing the switching losses,
maximum switching frequency of the converters is also another factor aimed at reducing the
as they are proportional to the switching frequency. The typical switching frequency for the power
switching losses, as they are proportional to the switching frequency. The typical switching frequency
converters is limited and typically lies between 20 kHz and 50 kHz. Additionally, there are considerable
for the power converters is limited and typically lies between 20 kHz and 50 kHz. Additionally, there
transient effects in the converters, due to the existence of stray capacitances and inductances in the
are considerable transient effects in the converters, due to the existence of stray capacitances and
power circuits, which lead to the rise in electromagnetic interference (EMI) problems [5].
inductances in the power circuits, which lead to the rise in electromagnetic interference (EMI)
Figure 11 presents a single-switch chopper. The switching device can be selected as a BJT, thyristor,
problems [5].
MOSFET or IGBT. Its operation is limited to the first quadrant, which means motoring operation, as it
Figure 11 presents a single-switch chopper. The switching device can be selected as a BJT,
can only supply positive voltage and current into the DC motor’s armature. In the case of continuous
thyristor, MOSFET or IGBT. Its operation is limited to the first quadrant, which means motoring
conduction operation, the equations which govern its operation are [5],
operation, as it can only supply positive voltage and current into the DC motor’s armature. In the
case of continuous conduction operation, theVequations which govern its operation are [5],
o = D·Vin , (7)
Vo = D·Vin , (7)
q
Vr = (1 −− DD)
VrV=inVin DD(1 ) ,, (8)
(8)
Vinin −
− eebb τ τ
Isw== D
Isw D −− Ip-p , (9)
R
Raa T Τ Ip-p , (9)
1−D
eb 1 − D
τ
IDI == τ IΙp-p − , , (10)
D T p-p − eb Ra (10)
Τ Ra
VV

in
Ra in1 1−− e−D·T/τ −(1−D)·T/τ
e D∙T/τ 1 1−−e e ( )∙ /
Ip-p = Ra , (11)
Ιp-p = 11−−e−T/τ ,
e /
La
τ τ== L,a , (12)
(12)
RR
aa
where VVoo,, eebb ,, V

where Vrr,, Isw
Isw, ,IDI,DΙ,p-p , T,, D
Ip-p T, are thethe
D are average
averageoutput
outputvoltage,
voltage,back
backEMF
EMF voltage,
voltage, output AC
output AC
ripple voltage, average switch current, average average diode
diode current,
current, peak
peak to
to peak
peak output
output ripple
ripple current,
current,
switching period, and duty cycle, respectively. respectively.
Figure 12a presents a two-quadrant hard-switching DC motor drive [5]. The first-quadrant
chopper is formed by the SW1 and D1 components through which the energy is fed by the Vin source
to the DC motor, resulting into the motoring mode of operation. Correspondingly, a fourth-quadrant
operation is formed by the SW2 and D2 components by which the energy is fed by the Vin source to the
DC motor, resulting into the regeneration mode of operation. Equations (7)–(12) are still valid, when
when the circuit operates in continuous conduction mode. The average diodes and switches currents
are dependent on the polarity of the maximum and minimum peak output current, and on the
polarity of the mean value of the output current. For these reasons, the estimation of these average
current values is more complex. When the converter operates in the first quadrant, the average
currents 2020,
Electronics flowing
9, 887through the SW1 and D1 components can be calculated using Equations (9) and (1),
10 of 26
respectively, while the corresponding currents flowing through the SW2 and D2 components are zero
[5].
the circuit
Figureoperates in continuous
12b depicts an H-bridge conduction mode. DC
or four-quadrant Thechopper
average [5].
diodes
The and switches
existence currents
of four are
switches
dependent on the polarity of the maximum and minimum peak output current, and
leads into different control methods in order to produce four-quadrant output voltage and current on the polarity of
thethe
by mean value of
meaning of the output current.
bidirectional For
current these
and reasons,
voltage. the estimation
Aiming at reducingofthe
these average in
distortion, current values
all methods
is more complex. When the converter operates in the first quadrant, the average currents
complementary switching components are used in each leg (T1 or T4 can be in “on” state but not both flowing
through
and the SW1 and T2
correspondingly D1and
components
T3 but notcanboth).
be calculated using Equations
The described operation (9) and (1), respectively,
philosophy ensures the
while
current the corresponding
Electronicscontinuity
2019, 8, x FORaroundcurrents flowing through the
the zero-current output [5].
PEER REVIEW SW2 and D2 components are zero [5]. 10 of 27
This H-bridge can be controlled in the same way such as, when two independent two-quadrant
choppers to be controlled. The main restriction is that the switches in the same leg must not be
SW
conducted simultaneously. The first chopper is formed by the conduction Rα of T1 and T4 switches with
D1 and D4 diodes. The first chopper works at first and second quadrants, where output current is
positive while output voltage is bidirectional, i.e. ±Vo. A second La
chopper is formed by the conduction
Vin D Vo
of T2 and T3 switches with D2 and D3 diodes, which means an operation at third and fourth
quadrants. In this case, a negative output current is available, with eb bidirectional output voltage ±Vo.
The two above-mentioned two-quadrant choppers, embedded in the same topology, ensure the DC
motor’s operation into the four quadrants. Another control method, for the H-bridge converter, is to
unify the operation of allFigure
four switches
Figure 11. [5].
11. One-quadrant
One-quadrant hard switching
hard switching DC
DC motor
motor drive.
drive.
Figure 12a presents a two-quadrant hard-switching DC motor drive [5]. The first-quadrant
chopper is formed by the SW1 SW1Dand D1 components through SW1 which D2
the energySW3is fedDby the Vin source
2
to the DC motor, resulting into the motoring mode of operation. Correspondingly, a 4fourth-quadrant
Rα La La
operation is formed by Vin the SW2 and D2 components eb by whichRαthe energy is fed by the Vin source to
Vin eb
the DC motor, resulting into the regeneration Vo mode of operation. Equations
Vo
(7)–(12) are still valid,
when the circuit operates SW2 in continuous
D1
conduction mode. The average diodes and switches currents
SW2 D1 SW4 D3
are dependent on the polarity of the maximum and minimum peak output current, and on the
polarity of the mean value of the output current. For these reasons, the estimation of these average
current values is more complex. When the converter operates in the first quadrant, the average
currents flowing through the SW1 and D1 components can be calculated using Equations (9) and (1),
Figure 12.
Figure (a) Two-quadrant
12. (a) Two-quadrant and (b) (b) Four-quadrant
Four-quadrant hard-switching DC
DC motor drives.
respectively, while the corresponding and currents flowing through hard-switching
the SW2 and motor drives.
D2 components are zero
[5]. Figure 12b depicts an H-bridge or four-quadrant DC chopper [5]. The existence of four switches
Depending on whether the zero output voltage DC loops are employed or not, using the above
leadsFigure 12b depicts
into different controlan methods
H-bridge ororder
in four-quadrant
to produce chopper
four-quadrant [5]. output
The existence
voltageofandfour switches
current by
control
leads methods,
into differentthe converter’s
control methods output voltage
in order can be
to produce either bipolar or multilevel. In the case of a
the meaning of bidirectional current and voltage. Aimingfour-quadrant
at reducing theoutput voltage
distortion, andmethods
in all current
by the meaning of
complementary bidirectional
switching currentare
components andused
voltage. Aiming
in each leg (T1 at reducing
or T4 can thebe indistortion,
“on” state inbut
all methods
not both
and correspondingly T2 and T3 but not both). The described operation philosophy ensures thenot
complementary switching components are used in each leg (T1 or T4 can be in “on” state but both
current
and correspondingly T2 and T3
continuity around the zero-current output [5]. but not both). The described operation philosophy ensures the
current
Thiscontinuity
H-bridgearound the zero-current
can be controlled output
in the same [5].such as, when two independent two-quadrant
way
This H-bridge can be controlled in
choppers to be controlled. The main restriction is thatthe same way such theas,switches
when two in independent
the same legtwo-quadrant
must not be
choppers to be controlled. The main restriction is that the switches
conducted simultaneously. The first chopper is formed by the conduction of T1 and T4 switches in the same leg must not be
with
conducted
D1 and D4 simultaneously.
diodes. The firstThe first chopper
chopper works at is formed
first andbysecond
the conduction
quadrants, of where
T1 andoutput
T4 switches
currentwith
is
D1 and D4 diodes. The first chopper works at first and second quadrants,
positive while output voltage is bidirectional, i.e. ±Vo. A second chopper is formed by the conduction where output current is
positive
of T2 andwhile output with
T3 switches voltage
D2 isand bidirectional,
D3 diodes, whichi.e. ±Vo. A second
means chopperatisthird
an operation formedandby the conduction
fourth quadrants.
In this case, a negative output current is available, with bidirectional output voltage ±Vand
of T2 and T3 switches with D2 and D3 diodes, which means an operation at third fourth
o . The two
quadrants. In this case, a negative output current is available, with
above-mentioned two-quadrant choppers, embedded in the same topology, ensure the DC motor’s bidirectional output voltage ±Vo.
The two above-mentioned
operation into the four quadrants. two-quadrant
Anotherchoppers, embedded
control method, in the
for the same topology,
H-bridge converter,ensure the DC
is to unify the
motor’s operation into the
operation of all four switches [5]. four quadrants. Another control method, for the H-bridge converter, is to
unifyDepending
the operationon of all fourthe
whether switches [5].
zero output voltage loops are employed or not, using the above
control methods, the converter’s output voltage can be either bipolar or multilevel. In the case of a
bipolar output, there is an increment to the ripple current value. However, a faster current reversal is
SW1 crossover
facilitated, with minimum D2 SW1 of the
distortion. The direction SW3
D2 output current D4 not influence
does
the operation. According to theRα above,
La two H-bridge control techniques
Rα La can be considered which are
eb
Vin output voltage and
known as the three-level the bipolar
Vin output voltage. eb
Vo Vo
SW2 D1 SW2 D1 SW4 D3
Electronics 2020, 9, 887 11 of 26
In the case of a bipolar output voltage, the average output voltage and the AC ripple voltage can
be described by the following equations [5]:
Vo = (2D − 1)Vin , (13)

q
Vr = 2Vin D(1 − D), (14)
The peak to peak output ripple current is twice the value given by (11). For three-level output
and for D ≤ 0.5:
Vo = (2D − 1)Vin , (15)
√ q
Vr = 2 Vin D(1 − 2D), (16)
and for D ≥ 0.5

Vo = (2D − 1)Vin , (17)
√ q
Vr = 2 Vin (2D − 1)(1 − D) (18)
2.2.2. Soft-Switching Converters for DC Drives

During the hard-switching operation of DC to DC converters, the electronic switches operate under
the full load when the switches turn on or turn off. This kind of operation stresses the semi-conductors
leading into increased losses. These losses are linearly increased with the converters’ switching
frequency. Additionally, the large dv/dt and di/dt result in increased EMI problems. The above
abnormal conditions are originated from the need of smaller converters’ size and weight, thus increased
power density.
A strategy to minimize the above-mentioned shortcomings is to change converter’s switch state
when the current flowing through it or the voltage across it, is zero at switching instant. This can be
succeeded with the usage of LC resonant circuits, aiming to shape the voltage or current waveforms
forcing the power device to switch at zero-voltage (ZV) or zero-current (ZC) condition. The usage of LC
resonant circuits leads to the well-known “resonant soft-switching” converters. In the literature [16–26],
soft-switching DC-DC topologies have been proposed for DC motor drives, offering the ability for
motoring and regenerative braking, an operation which can be achieved by the bidirectional power flow.
The operation of soft-switching topologies is based on resonant elements, while their characteristic
impedance and their angular frequency ω can be described as:
p p
Z = Lr /Cr , ω = 1/Lr Cr (19)
Figure 13 illustrates a two-quadrant (2Q) Zero-Voltage Multi-Resonant (ZVMR) converter [18].

This topology is composed by a conventional 2Q-PWM DC drive, two resonant capacitors and a
resonant inductor. It is applied in systems where motoring and regenerative braking modes of DC
motors are to be achieved, offering by this way bidirectional flow in electric power. One advantage for
this soft-switching converter is the operation with high switching frequencies, typical above 100 kHz,
with the ripple current value of DC motor maintained in low magnitudes, while the switching losses are
quite low due to Zero-Voltage-Switching (ZVS) for all switches. Additionally, this converter provides
load variation and full ranges of voltage conversion.
This ZVMR converter uses all the built-in diodes and has the ability to absorb all major parasitic.
Additionally, it is important to be noted that the usage of ZVR technology is highly desirable for
converters using MOSFETs. This is because MOSFETs suffer from extremely high magnitudes in
capacitive voltage turn-on losses. As the 2Q-ZVMR behaves as a constant source after reaching the
maximum output current value, it handles short-circuit down to no-load conditions without any
additional measures.
maximum
Electronics output
2020, current value, it handles short-circuit down to no-load conditions without
9, 887 any
12 of 26
additional measures.
Figure 13. Zero-voltage multi-resonant converter fed DC motor drive.
The power rating of the semiconductors (MOSFET) associated with the MR cell is higher as
compared with the conventional two-quadrant PWM DC drive, due to the circulating energy and the
conduction losses.
Figure 14 illustrates a 2Q Zero-Voltage-Transition (ZVT) converter [19], and it has the ability to
operate in both motoring and
Figure 13. regenerative
Zero-voltage braking modes
multi-resonant of DC
converter fedmotor drives,
DC motor by this way offering
drive.
Zero-voltage
bidirectional flow in electric power. In contrast to the conventional 2Q-PWM DC drives, the 2Q-ZVT
converter
The needs some
The power
power ratingextra
rating of
of the components
the such as,
semiconductors
semiconductors a resonant
(MOSFET)
(MOSFET) capacitor,with
associated
associated two the
with auxiliary
the MR cell
MR switches
cell is higher
is andas
higher asa
resonant
comparedinductor.
compared with
with the conventional two-quadrant
the conventional two-quadrant PWM PWM DC DC drive,
drive, due to the
due to the circulating
circulating energy
energy and and the
the
The
conduction 2Q-ZVT
losses.
conduction losses. converter presents some basic advantages such as unity device current and voltage
stressFigure
during
Figure 14the
14 motoring
illustrates
illustrates aa 2Qand
2Q regenerative modes of (ZVT)
Zero-Voltage-Transition
Zero-Voltage-Transition operation,
(ZVT) zero voltage
converter
converter [19],
[19], and
andswitching
it
it has thefor
has the all the
ability
ability to
to
diodes
operate and
in main
both switches,
motoring andsame resonant
regenerative tank
brakingfor both,
modes backward
of DC and
motor forward
drives,
operate in both motoring and regenerative braking modes of DC motor drives, by this way offering by power
this way flows, full
offering
utilization
bidirectional
bidirectional forflow
all the
flow in built-inpower.
in electric
electric diodesIn
power. Inofcontrast
the power
contrast switches
to the
to the and simple
conventional
conventional 2Q-PWM
2Q-PWMcircuitDCtopology.
DC drives, the
drives, The
the above
2Q-ZVT
2Q-ZVT
advantages
converter needs
converter make
needs some the 2Q-ZVT
some extra converters
extra components
components such to operate
such as, with high
as, aa resonant switching
resonant capacitor,
capacitor, two frequency,
two auxiliary high
auxiliary switches efficiency
switches and and a a
and high-power
resonant inductor.
resonant inductor. density.
The 2Q-ZVT converter presents some basic advantages such as unity device current and voltage
stress during the motoring and regenerative modes of operation, zero voltage switching for all the
diodes and main switches, same resonant tank for both, backward and forward power flows, full
utilization for all the built-in diodes of the power switches and simple circuit
Rα topology. The above
advantages make the 2Q-ZVT converters to operate with high switching frequency, high efficiency
and high-power density. Vo La
Vin
eb
α R
Figure
Figure 14.
14. Zero-voltage-transition
Zero-voltage-transition converter
converter fed
fed DC
DC motor
motor drive.
drive.
Vo La
Additionally,
The 2Q-ZVT converteran appropriate
presents control
somesystem which will such
basic advantages control the
as unity semiconductor
device currentswitches is of
and voltage
Vin
major importance
stress during for this converter
the motoring type. Thismodes
and regenerative controlofsystem can be
operation, achieved
zero voltageusing a Digital
switching Signal
for all the
e
Processor
diodes and main switches, same resonant tank for both, backward andb forward power flows, full
(DSP).
Figurefor
utilization 15 all
depicts a 2Q Zero-Current-Transition
the built-in diodes of the power switches (ZCT) and
converter
simple[21]. As topology.
circuit for the twoThe previous
above
topologies,
advantages make this converter
the 2Q-ZVT canconverters
operate intomotoring andhigh
operate with regenerative
switching braking
frequency, operation modes.and
high efficiency In
Figure 14. Zero-voltage-transition converter fed DC motor drive.
comparison to its PWM counterpart, this topology requires a resonant capacitor, a resonant inductor
high-power density.
and two auxiliary switches.
Additionally,
Additionally, an
an appropriate
appropriate control
control system
system which
which will
will control
control the
the semiconductor
semiconductor switches
switches is is of
of
major This converter
importance fortype
this exhibits
converter some
type.important
This advantage
control system such
can as
be the
achievedzero-current
major importance for this converter type. This control system can be achieved using a Digital Signal using a switching
Digital for
Signal
all the diodes
Processor
Processor (DSP).
(DSP).and switches, main and auxiliary, same resonant tan for both backward and forward
power flows,
Figure
Figure 15 full
15 utilization
depicts
depicts aa 2Q of all built-in diodes of(ZCT)
2Q Zero-Current-Transition
Zero-Current-Transition the power
(ZCT) switches,
converter
converter [21]. As
[21]. As minimum
for the
for current
the two
two and
previous
previous
voltage
topologies,stresses,
this low cost
converter and
can simple circuit
operate in topology.
motoring These
and characteristics
regenerative
topologies, this converter can operate in motoring and regenerative braking operation modes. In lead
braking to the deployment
operation modes.
In comparison
comparison toPWM
to its its PWM counterpart,
counterpart, this topology
this topology requires
requires a resonant
a resonant capacitor,capacitor,
a resonant a resonant
inductor
inductor and two auxiliary
and two auxiliary switches. switches.
This converter
This converter typetype exhibits
exhibits some
some important advantage such
important advantage such as
as the
the zero-current
zero-current switching
switching for for
all the
all the diodes
diodes and and switches,
switches, mainmain andand auxiliary,
auxiliary, same
same resonant
resonant tan
tan for
for both
both backward
backward and and forward
forward
power flows, full utilization of all built-in diodes of the power switches,
power flows, full utilization of all built-in diodes of the power switches, minimum currentminimum current and voltage
and
stresses, low cost and simple circuit topology. These characteristics lead to the deployment
voltage stresses, low cost and simple circuit topology. These characteristics lead to the deployment of medium
frequency switching characteristics, in the range of 50 kHz, high efficiency and high-power density.
of medium frequency switching characteristics, in the range of 50 kHz, high efficiency and high-
power
of medium density.
Electronics 2020,frequency
9, 887 switching characteristics, in the range of 50 kHz, high efficiency and13high-
of 26
power density.
Rα
Rα
Vin Vo La
Vin Vo La
eb
eb
Figure 15. Zero-current-transition converter fed DC motor drive.

Figure 15. Zero-current-transition
Figure 15. Zero-current-transition converter
converter fed
fed DC
DC motor
motor drive.
drive.
The 2Q-ZCT converter is usually applied for medium-power DC motor applications, in the
rangeThe
The of2Q-ZCT
few kW,
2Q-ZCT converter
using as
converter is is
usually
main
usually applied
switching
applied forpower
medium-power
for devices the
medium-power DC motor
IGBTs
DC motor applications,
(Insulated-Gate
applications,in theBipolar
range
in the
of few kW,
Transistors).
range of few using
IGBTs
kW, asusually
main switching
using suffer
as main frompower
severe
switching devices
inductive
power thedevices
IGBTs
turn-off(Insulated-Gate
andIGBTs
the diode reverse Bipolar
recovery
(Insulated-Gate Transistors).
turn-on
Bipolar
IGBTs
losses. usuallyIGBTs
Transistors). sufferusually
from severesufferinductive
from severe turn-off and diode
inductive turn-off reverse recovery
and diode turn-on
reverse losses.turn-on
recovery
Figure
Figure 16 presents a four-quadrant Zero-Voltage-Transition (4Q-ZVT) converterMOSFET
losses. 16 presents a four-quadrant Zero-Voltage-Transition (4Q-ZVT) converter [21,22]. [21,22].
power
MOSFET semiconductors
Figure power16 presents are amainly
semiconductors used
are in this
mainly
four-quadrant usedtypeinofthis converter. It also (4Q-ZVT)
type of converter.
Zero-Voltage-Transition gives the chance
It also gives
converter of operation
the chance[21,22].of
in motoring
operation in and regenerative
motoring and braking
regenerative in both
braking forwardin and
both reversible
forward
MOSFET power semiconductors are mainly used in this type of converter. It also gives the chance of operation
and reversible of DC motor
operation drives.
of DC
The
motor4Q-ZVT
operation inexhibits
drives. The 4Q-ZVT
motoring someandbasic advantages
exhibits
regenerative such as
somebraking
basic insimple
both circuit
advantages suchtopology,
forward asand
simple unity current
circuit
reversible and voltage
topology,
operation ofunity
DC
stress
motorand
current and zero
drives. voltage
voltage
The 4Q-ZVTswitching
stress and zerofor
exhibits all main
voltage
some and
switching
basic auxiliary diodes
for all main
advantages suchand asand switches.
auxiliary
simple Additionally,
diodes
circuit and switches.
topology, full
unity
utilization
current andofvoltage
Additionally, all
fullbuilt-in
stressdiodes
utilization and of built-in
ofzero
all the power
voltage diodes switches
switching of the can
for allbe
power mainachieved
switches minimizing
can be achieved
and auxiliary diodes byand this way the
minimizing
switches.
overall
by this hardware
way the
Additionally, cost, and
fulloverall
utilization itofuses
hardware the same
cost,
all built-in and itresonant
diodes usesof thetanksame
power forresonant
both backward
switches tank
can be and
bothforward
forachieved backward power
minimizing and
flows.
by thisThe
forward way above-mentioned
power theflows.
overall The characteristics
cost, andlead
above-mentioned
hardware to highthe efficiency
characteristics
it uses same lead toandhigh
resonant power
tank density
efficiencyfor bothandofbackward
the converter.
power density
and
The
of zero-voltage
the converter.
forward power flows. switching
The zero-voltage operation can be realized
switching operation
The above-mentioned with
characteristics the
can be utilization
leadrealized
to highwith of two sets of
the utilization
efficiency resonant
and power of two tanks.
sets
density
The first
of resonant
the set includes,
tanks.The
converter. Thethe resonant
first capacitors
set includes,
zero-voltage a /2, thecapacitors
Coperation
the resonant
switching inductor LCa aand
can be realized the
/2, the auxiliary
inductor
with Lswitches
a and the
the utilization ofSauxiliary
andsets
atwo Sa ’
for soft
switches switching
S and
of resonant tanks. The
a S S
a ’ 1 and
for S
soft . The second
switching S set
and includes,
S . The the
second resonant
set
first4 set includes, the resonant capacitors Ca/2, the inductor
1 4 capacitors
includes, the C /2,
resonant
b the inductor
capacitors
La and the auxiliary L and
b /2,
C b
the auxiliary
inductor
switches Sa and switches
Lb and
Sa’ for S
thesoft and
bauxiliaryS ’ for soft
b switches
switching switching
S1 andSbS4and S
. TheSbsecond
’ forand
2 soft S
set . As
3switchingit can
includes,Sthe clearly
2 and be seen in
S3. As itcapacitors
resonant Figure
can clearly Cb16,
be
/2,
there
seen areFigure
in
the inductor two L2Q-ZVT
b16,
and there
theconverters
are
auxiliary which
two 2Q-ZVT
switches simultaneously
converters
Sb and Sbwhich ’ forfed the DC motor.
simultaneously
soft switching fed Sthe
S2 and DCitmotor.
3. As can clearly be
seen in Figure 16, there are two 2Q-ZVT converters which simultaneously fed the DC motor.
Sα S1 S3 Sb′
Dα D1 Ca / 2 D3 Cb / 2 Db′
Sα S1 S3 Sb′
DαLa D1 Ca / 2 D3 Cb /L2 Db′
Rα L b
Vin eb
La Rα L Lb
Vo eb
SVαin′ S4 S2 Sb
Dα ′ D4 Ca / 2 Vo D2 Cb / 2 Db
Sα ′ S4 S2 Sb
Dα ′ D4 Ca / 2 D2 Cb / 2 Db
Figure 16. 4Q-ZVT converter fed DC motor

Figure 16. motor drive.
drive.
Figure 16. 4Q-ZVT converter fed DC motor drive.
Figure
Figure 17 17 presents
presents a a4Q-ZCT4Q-ZCT (four-quadrant
(four-quadrant Zero-Current-Transition)
Zero-Current-Transition) converter
converter [22]. [22].
The
The structuring
structuring power
power elements
elements for for this
this converter
converter areare
thethe IGBTs,
IGBTs, and
and
Figure 17 presents a 4Q-ZCT (four-quadrant Zero-Current-Transition) converter [22]. Theitit can
can bebe used
used for
for motoring
motoring and
and
regenerative
regenerative
structuring power braking
braking for DC
for
elements DCmotors,
motors,
for offeringthe
this offering
converter the possibility
possibility
are the IGBTs, for
for forward
forward
and it canand and
be reversible
reversible
used operations
operations
for motoring as
and
as
well.well.
regenerative braking for DC motors, offering the possibility for forward and reversible operations as
well.The
The achievement
achievement of of ZCS
ZCS operation
operation requires
requires twotwo sets
sets of of resonant
resonant tanks.tanks. TheThe first
first tank
tank includes
includes aa
resonant
resonant capacitor
The capacitor
achievementC , an
Caa, an inductor
inductor
of ZCS L and the
La aandrequires
operation auxiliary
the auxiliary
two sets switches
switches S
Sa and
of resonant and S ’,
Sa’, for
a tanks. a Thefor soft
softfirst switching
switching of S1ofand
tank includes S1a
and S4 . second
Sresonant
4. The The second
one C
capacitor one includes
includes a resonant
a resonant
a, an inductor La and capacitor
capacitor
the Cbswitches
Cb, an
auxiliary , an inductor
inductor SL LbSaand
b and
a and the the auxiliary
auxiliary
’, for soft switches
switches
switching Sb
of Sb1 and
and S
Sb4.’ The ’ in
in order
b order to achieve
to achieve
second soft switching
soft switching
one includes a resonant of
ofcapacitorS and
S2 and S3.CAs
2 S .
it can
b, an3 As it can clearly
clearlyLbbe
inductor and seen be seen
theinauxiliary in
Figure 17,Figure 17,
there are
switches there
Sb two
and
are
2Q-ZCTtwo
Sb’ in 2Q-ZCT
order converters
converters
to achieve which which
soft switching simultaneously
simultaneously
of S2 andfed feditDC
the
S3. As the
canDC motor.
motor.
clearly The
be The
seen 4Q-ZCT
4Q-ZCT
in Figure presents
presents
17, there the
thearesame
same
two
characteristics
2Q-ZCT converters as its 2Q-ZCT
which ancestor.
simultaneouslyThe 4Q-ZCT
fed the topology
DC motor. is applicable
The 4Q-ZCT for DC motors
presentswiththepower
same
up to 5kW, while its semiconductors can operate in high frequencies.
characteristics
Electronics 2020, 9, as
887its 2Q-ZCT ancestor. The 4Q-ZCT topology is applicable for DC motors with power
14 of 26
up to 5kW, while its semiconductors can operate in high frequencies.
Sα ′ S1 S3 Sb
Dα ′ D1 D3 Db
La Ca Rα L Cb Lb
Vin eb
Vo
Sα S4 S2 Sb′
Dα D4 D2 Db′
Figure 17. 4Q-ZCT converter fed DC motor drive.
2.3. Comparative Evaluation of DC Motor Drives

2.3. Comparative Evaluation of DC Motor Drives
In industrial applications, such as printing, mining, cranes and metals shaping applications,
In industrial applications, such as printing, mining, cranes and metals shaping applications,
controlled rectifier-fed DC drives are common choice. If the converter is indented to supply motors up
controlled rectifier-fed DC drives are common choice. If the converter is indented to supply motors
to a few kilowatts then both the three-phase and single-phase mains could be used. However, when
up to a few kilowatts then both the three-phase and single-phase mains could be used. However,
the application includes larger motors, with power above the 15 kW then traditional three-phase is
when the application includes larger motors, with power above the 15 kW then traditional three-
used as main supply. The standard compact converters are available in sizes from 0.5 kW up to several
phase is used as main supply. The standard compact converters are available in sizes from 0.5 kW up
hundred kW. The controlled rectifier circuits with are available are differentiated according to the
to several hundred kW. The controlled rectifier circuits with are available are differentiated according
application. The single-phase controlled rectifiers can be categorized into one-quadrant, two-quadrant,
to the application. The single-phase controlled rectifiers can be categorized into one-quadrant, two-
and four-quadrant topologies.
quadrant, and four-quadrant topologies.
In the case of one-quadrant converters, both half wave topologies and asymmetrical
In the case of one-quadrant converters, both half wave topologies and asymmetrical semi-
semi-converters have been manufactured. The half wave 1Q topologies are used to control DC
converters have been manufactured. The half wave 1Q topologies are used to control DC motors with
motors with power up to 0.5 kW, while the asymmetrical one is used for power up to 15 kW. The
power up to 0.5 kW, while the asymmetrical one is used for power up to 15 kW. The half-controlled
half-controlled converters present important characteristics, in comparison to the fully controlled, such
converters present important characteristics, in comparison to the fully controlled, such as cheaper
as cheaper control circuits, better power factor, increased mean value of the output voltage for the
control circuits, better power factor, increased mean value of the output voltage for the same firing
same firing angle, and smaller total harmonic distortion on the current in the input side. Moreover, in
angle, and smaller total harmonic distortion on the current in the input side. Moreover, in the case of
the case of 2Q topologies, the full wave converters are more appropriate for applications up to 15 kW,
2Q topologies, the full wave converters are more appropriate for applications up to 15 kW, while if a
while if a 4Q operation is of major need then the single-phase dual converter is the ideal choice for
4Q operation is of major need then the single-phase dual converter is the ideal choice for power up
power up to 15 kW.
to 15 kW.
Three-phase controlled rectifiers can also be categorized in 1Q, 2Q and 4Q topologies. The most
Three-phase controlled rectifiers can also be categorized in 1Q, 2Q and 4Q topologies. The most
famous and traditionally used AC to DC converter is the fully controlled, especially for applications
famous and traditionally used AC to DC converter is the fully controlled, especially for applications
with rated power up to 140 kW, although there are half-controlled three-phase converters. In the case
with rated power up to 140 kW, although there are half-controlled three-phase converters. In the case
of half-controlled topology (three diodes replace the three thyristors) total harmonic distortion on the
of half-controlled topology (three diodes replace the three thyristors) total harmonic distortion on the
line current waveform is increased the and consequently it is assumed to be unsuitable especially for
line current waveform is increased the and consequently it is assumed to be unsuitable especially for
large applications. Furthermore, the half-controlled converters provide 1Q operation in contrast to the
large applications. Furthermore, the half-controlled converters provide 1Q operation in contrast to
fully-controlled which provide 2Q operation. Finally, when the 4Q operation is required, then the dual
the fully-controlled which provide 2Q operation. Finally, when the 4Q operation is required, then the
three-phase fully-controlled converters should be used.
dual three-phase fully-controlled converters should be used.
On the other hand, choppers are used worldwide especially for rapid transit systems. Because of
On the other hand, choppers are used worldwide especially for rapid transit systems. Because
their high efficiency, fast response and the regeneration capability, they have replaced the conventional
of their high efficiency, fast response and the regeneration capability, they have replaced the
controlled-rectifier converter in many DC applications. Additionally, they present decreased motor
conventional controlled-rectifier converter in many DC applications. Additionally, they present
losses and torque ripple, because of the decreased armature ripple current, a phenomenon resulting
decreased motor losses and torque ripple, because of the decreased armature ripple current, a
from the high switching frequency.
phenomenon resulting from the high switching frequency.
Choppers can also be divided in two major categories: Hard-switching and soft-switching. These
Choppers can also be divided in two major categories: Hard-switching and soft-switching. These
converters can both operate in one, two or four quadrants. Depending on the required frequency and
converters can both operate in one, two or four quadrants. Depending on the required frequency and
the DC motor power, hard-switching choppers utilize GTOs, BJTs, MCTs, MOSFETs, thyristors and
the DC motor power, hard-switching choppers utilize GTOs, BJTs, MCTs, MOSFETs, thyristors and
IGBTs as semiconductor switches. In case that low switching frequencies, up to several hundred Hz,
IGBTs as semiconductor switches. In case that low switching frequencies, up to several hundred Hz,
and high-power requirements are needed, thyristors and GTOs are utilized. When the application
and high-power requirements are needed, thyristors and GTOs are utilized. When the application
requires high switching frequencies, in the range of 20–50 kHz, and the motors’ power is limited to low
requires high switching frequencies, in the range of 20–50 kHz, and the motors’ power is limited to
or medium values, then IGBTs and MOSFETs are mainly used, and in some cases BJTs as well. The
employment of PWM control techniques is usually needed for all the above chopper cases.
Electronics 2020, 9, 887 15 of 26
When the application requires higher switching frequencies, the use of soft-switching choppers
is the ideal choice, aiming to limit the Electromagnetic Interference (EMI) and to reduce switching
losses. Such applications include the low-inductance DC motors. Moreover, the use of soft-switching
techniques can increase the converters’ efficiency by reducing the converters’ losses.
In the case of a 2Q-ZVMR converter, power switches can operate with high frequencies, above
100 kHz, with low value ripple current, while the switching losses tend to be zero due to the zero-voltage
switching. Additionally, this converter type can operate with load variation and provides full ranges
of voltage conversion. In contrast to the conventional 2Q-PWM DC drives, the conduction losses and
circulating energy lead to higher power rating of the MOSFET associated with the MR cell.
When MOSFETs are to be used then a 2Q-ZVT converter is proposed in contrast to the ZVMR.
In case which IGBTs are chosen to be used as power switches, then the 2Q-ZCT converter should
be selected instead of the ZVMR. The above selections are proposed because of the advantages that
these converters present, such as the unity device current and voltage stress during both regenerative
and motoring modes, the same resonant tank for both backward and forwards power flows, the zero
voltage switching for all main diodes and switches, the full utilization of all built-in diodes of the
power switches, and the simple circuit topology. Due to the above advantages, high efficiency and
power density at high switching frequencies (>100 kHz) can be achieved. The use of 2Q-ZCT converter
is applied on medium power DC motors, utilizing IGBTs which operate under switching frequencies
in the range of 50 kHz. The proper driving of the semiconductors utilized in the above soft switching
converter types, impose the application of a DSP system.
The 4Q soft switching converters (4Q-ZVT and 4Q-ZCT) are also applied with MOSFETs and
IGBTs as power switches. Their usage is proposed for applications with rated power up to about 5 kW,
with high switching frequency, while the switching driving signals are traditionally also produced by a
DSP system.
Finally, the summary of the key characteristics of chopper-fed drives are presented in Table 2.
Table 2. Main characteristics of chopper-fed drives.
General Features
• High order harmonics • 0.5 kW to

• Reduced ripple current • High frequency operation (∝ kHz)
• Suffer from excessive voltage and hundred kW
• Regeneration operation • High response speed
current stresses • High Efficiency
Hard-Switching Converters
• Typical switching frequency • Considerable transient effects

• High switching stress and losses
between 20 kHz and 50 kHz in the converters
H-Bridge or 4Q Chopper
Single-switch Chopper
• 4Q Operation
• 1Q operation • Increased number of switches
• Simple control circuit • Medium complexity control circuit
2Q Chopper • Current continuity around the
• 2Q operation zero-current output
• Increased number of switches • Multilevel output voltage
• Medium complexity control circuit • Bipolar output voltage => increment to the
ripple current value, faster reversal, with
minimum crossover distortion
Electronics 2020, 9, 887 16 of 26
Table 2. Cont.
Soft-Switching Converters
• Usage of LC resonant circuits

• High switching frequency, above to 100 kHz • Very High Efficiency
• Increased number of switches
• Low ElectroMagnetic Interference (EMI) • Complex Control Circuit
• Very low ripple current
2Q-ZVMR Converter
• 2Q Operation
• Quite low switching losses 4Q ZVT Converter
• Provides load variation and full ranges of voltage conversion • 4Q Operation
2Q ZVT Converter • Unity current and voltage stress
• 2Q Operation • Zero voltage switching
• Unity device current and voltage stress • High power density
• Zero voltage switching 4Q ZCT Converter
2Q ZCT Converter • 4Q Operation
• 2Q Operation • Same characteristics as 2Q-ZCT
• Zero current switching • Power up to 5 kW
• Minimum current and voltage stresses
• Switching frequency in the range of 50 kHz
• For medium-power DC motor applications in the range of few kW
3. Control Techniques for DC Motor Drives

During the last decades, a lot of different control techniques have been introduced in order to
control the speed and the torque of direct current motors. The control techniques are very important in
order to ensure proper function and performance of an electric drive system. In the bibliography there
are a lot of different control techniques and each technique can be applied with some different modes.
This review paper illustrates some basic control techniques based on conventional PID controllers and
intelligent controllers as well.
3.1. Classic PID Controllers

Over the last years, traditional controllers for DC motor were based on the Proportional Integrative
and Differential (PID) controller. Some other subcategories of DC motor controllers can occur, beginning
from the PID controller [23]:
â Proportional (P) based control technique

â Proportional–Integral (PI) based control technique
â Proportional–Differential (PD) based control technique
â Proportional–Integral–Differential (PID) based control technique
The idea of a PID controller and its subcategories is to take as a feedback a special characteristic of
the driving system and a setting point for this special characteristic, and applying some mathematical
processes to achieve this setting point to the system’s output. Such a special characteristic can be the
speed or/and the torque of the DC motor. The basic block diagram of a PID controller is illustrated in
Figure 18 where the PID controller has got as input the difference between feedback signal y(t) and
reference signal r(t) and this difference feeds the PID controller block [24].
As Figure 18 illustrates, the full PID controller consists of three control parts, the proportional
part, the integral part and the differential part. Each part calculates a specific mathematical operation.
These mathematical operations are shown in Figure 18 as well. The output of the full PID controller,
according to the Figure 18, is:
Zt
de(t)
u(t)= kp ·e(t)+ki e(τ)dτ + kd . (20)
dt
0
occur, beginning from the PID controller [23]:
 Proportional (P) based control technique
 Proportional–Integral (PI) based control technique
 Proportional–Differential (PD) based control technique
Electronics 2020, 9, 887 17 of 26
 Proportional–Integral–Differential (PID) based control technique
The idea of a PID controller and its subcategories is to take as a feedback a special characteristic
Whereas, the transfer function of the PID, calculated by the Laplace transform of (20) is,
of the driving system and a setting point for this special characteristic, and applying some
mathematical processes to achieve this usetting (t) point
ki to the system’s output. Such a special
characteristic can be the speed or/and G(sthe
) = torque = of + DC
kpthe +kd ·s The basic block diagram of a(21)
e(t) s motor. PID
controller is illustrated in Figure 18 where the PID controller has got as input the difference between
where kp , ksignal
feedback i , kd are
y(t)the gain
and factors signal
reference for proportional,
r(t) and thisintegral, andfeeds
difference derivative,
the PIDoperations
controllerrespectively.
block [24].
Figure18.
Figure 18.PID
PIDController
Controllerinternal
internalblock
blockstructure
structure[24].
[24].
Depending
As Figure on 18 the values of
illustrates, these
the full factors, the above
PID controller mentioned
consists different
of three conventional
control controllers
parts, the proportional
can be implemented. For example, if k
part, the integral part and the differential 0, k
p part. Each
, i ,, kpart
d , 0 then a PID controller is implemented. If
calculates a specific mathematical operation.
kpThese ki ,, kd = 0 then
, 0, mathematical a PI controller
operations are shownis implemented
in Figure 18 and so on.
as well. The output of the full PID controller,
In a PI the ‘P’ component
according to the Figure 18, is: determines the reaction to current error, whereas ‘I’ determines the
reaction to the sum of recently appeared errors.t Accordingde(t) to the PI tuning procedure, the typical
objectives are [25]: u(t) = kp ∙e(t) + ki e(τ)dτ + kd . (20)
0 dt
â Closed-loop
Whereas, thestability. The closed-loop
transfer function system
of the PID, outputby
calculated remains bounded
the Laplace for bounded
transform of (20)input.
is,
â Adequate performance. The closed-loop u(t) ki
system tracks reference changes and suppresses
Electronics 2019, 8, x FOR PEER REVIEW G(s) = = kp + + kd ∙s 18 of (21)
27
disturbances as rapidly as possible.e(t) The larger s the loop bandwidth, the faster the controller
where 𝑘 , 𝑘 ,to
responds 𝑘 changes
are the gain
inThe
the factorsorfor
reference proportional,
disturbances in the integral,
loop. andand derivative, operations
 Adequate robustness. loop design has enough gain margin phase margin to allow for
respectively.
â modeling
Adequateerrors robustness. The loop design dynamics.
has enough gain margin and phase margin to allow for
or variations in system
Depending
modeling on the
errors or values of these
variations factors,
in system the above mentioned different conventional controllers
dynamics.
canFigure 18 is the basic
be implemented. For operational
example, if PID 𝑘 ≠block
0, 𝑘 ≠, 𝑘 ≠ 0 but
diagram thenitacan PIDbecontroller
optimized is using an extra If
implemented.
𝑘 ≠ Figure
internal 0,PID
𝑘 ≠, 18 𝑘is the
loop =in0basic
the operational
block
then a PIdiagram PID
controller block
in is
order diagram
to control
implemented butthe
and it can be optimized
armature
so on. currentusing
in caseanofextra internal
a failure as
PID suggests.
[26] loop in the
The block
idea diagram
behind in
this order to
internal control
loop is the
the armature
case at current
which a in
very
In a PI the ‘P’ component determines the reaction to current error, whereas ‘I’ determines the case
large of a
load,failure
bigger as [26]
than
suggests.
the nominal
reaction The
to load
theideaof behind
sum theofDC this internal
motor,
recently loop
is connected
appeared is through
errors.theAccording
case the
at which
motor’s
to thea very
axle. large
In thatload,
PI tuning case,bigger than
the motor
procedure, the the
will
typical
nominal
decrease
objectives load
its are of
speed the DC motor, is connected through the motor’s axle. In that
and the basic PID controller will increase the armature current, in order to drive
[25]: case, the motor will
decrease
the DC motorits speed
in aand the basic
specific speed PIDand controller will increase
torque output. the armature
The result current, in order
is the overloading of thetoDCdrive the
motor
DC motorClosed-loop
in a specificstability.
speed The torque
and closed-loop
output.system
The output
result is remains
the bounded
overloading offor
thebounded
DC motorinput.
and its
and its damage. Figure 19 illustrates the new PI block diagram using an internal armature PID
 Adequate
damage. Figure 19performance.
illustrates the The
new closed-loop
PI block diagramsystemusingtracks
an reference
internal changes
armature PID and suppresses
controller loop.
controller loop.
disturbances as rapidly as possible. The larger the loop bandwidth, the faster the controller
responds to changes in the reference or disturbances in the loop.
Figure
Figure 19.
19. Speed
Speedcontroller
controller two-quadrant
two-quadrant dc
dc motor
motor drive
drive [26].
[26].
There are a lot of applications using PID controller, in order to control DC or AC motors. Some
of these applications use extra techniques and different technical approaches to drive a motor system
[26,29,30]. In [26], as already mentioned, the usage of an extra internal PID controller is implemented
in order to control armature current against overloading. In [29] another PI controller is implemented
Electronics 2020, 9, 887 18 of 26
There are a lot of applications using PID controller, in order to control DC or AC motors.
Some of these applications use extra techniques and different technical approaches to drive a motor
system [26,29,30]. In [26], as already mentioned, the usage of an extra internal PID controller is
implemented in order to control armature current against overloading. In [29] another PI controller
is implemented for a golf terrain. In this implementation, the reference input is given through the
user. There is also a battery system and a buck converter used to ensure a type of autonomy for this
system. The DC motor can be a Permanent Magnet DC motor (PMDC) and the implementation of
PID controller is as simple as in a usual DC motor. Such an application is described in [30] where an
extra hardware implementation technique uses a Field Programmable Gate Array (FPGA) in order to
implement the PI controller. The usage of FPGA is a solution that avoids the high initial cost [31,32]
and the FPGA programmability permits design upgrades, with no hardware replacement necessary,
which is impossible with classic Application—Specific Integrated Circuits (ASICs).
The next step of a controller system is the production of pulses to drive the power electronic
topology, which is selected for the specific driving system. The most usual and dominant
pulse production technique, for controlling the analog circuits using digital signals, is the Pulse
Width Modulation (PWM) technique [33–35]. In [30], the PWM technique is implemented inside
the same FPGA chip with the PI controller. This means an extra saving cost and a simpler
hardware implementation.
3.2. Intelligent Controllers

Intelligent controllers are a particular category of controllers which use some special techniques
in order to drive DC or AC motors. The idea behind an intelligent controller is the pre-schedule
control procedure using a mathematical model of the system or a rule-based control procedure. For the
proper design of a motor drive system, the motor’s model, or in other words, its transfer function is
needed. This transfer function is derived by applying basic electrical and motion equations, given the
motor’s parameters. It is not unusual either for model simplicity or because of inaccurate knowledge
of the model itself, to model the motor with high amount of uncertainty in its transfer function. The
intelligent control techniques attempts to address, overcome and finally solve. There are a lot of
intelligent controllers such as Fuzzy Logic Controllers (FLC) and Artificial Neural Network Controllers
(ANNC) and some of these intelligent controllers are illustrated below.
3.2.1. Fuzzy Logic Controllers

Fuzzy Logic Controllers (FLCs) use the so-called Fuzzy Logic Technique (FLT) in order to
adequately control the drive system. The implementation of a FLC can be done as an individual part
to produce the control signals for the PWM or in combination with a PID controller. Sen et al. [36] have
included aspects of the fuzzy set theory in the control of DC motors back in 1996.
The basic constraints for DC motor drives are the non-linearity in dc motor, the variable and
unpredictable inputs, the noise propagation along a series of unit processes, the unknown parameters
and the changes in load dynamics [37]. The advantages of using FLT are the inherent approximation
capability, the high degree of tolerance, the smooth operation, the reduction of the effect of non-linearity,
the fast adaptation and the learning ability [37].
A Fuzzy Logic controller implements an if-then programming form for the various working
conditions to control the motor speed. The Fuzzy logic control is a control algorithm based on a
linguistic control strategy [38]. Crispy logic in Boolean theory is based on certain and precise situations
in contrast to the Fuzzy logic. The variables in fuzzy logic have set of values which are expressed by
linguistic expressions, such as SMALL, MEDIUM and LARGE [39]. These linguistic expressions are
represented numerically by fuzzy sets. Every fuzzy set is characterized by a membership function,
which varies between 0 and 1 in contrast to the Boolean sets, which can take values 0 or 1 [38]. A given
element can be a member of a fuzzy set with a degree varying from 0 to 1. The basic structure of a FLC
is illustrated in Figure 20. The basic Fuzzy Controller parts are [35,36,40]:
linguistic control strategy [38]. Crispy logic in Boolean theory is based on certain and precise
situations in contrast to the Fuzzy logic. The variables in fuzzy logic have set of values which are
expressed by linguistic expressions, such as SMALL, MEDIUM and LARGE [39]. These linguistic
expressions are represented numerically by fuzzy sets. Every fuzzy set is characterized by a
Electronics
membership 2020, 9, 887
function, which varies between 0 and 1 in contrast to the Boolean sets, which can 19 take
of 26
values 0 or 1 [38]. A given element can be a member of a fuzzy set with a degree varying from 0 to 1.
The basic
â structure of a FLC is illustrated in Figure 20. The basic Fuzzy Controller parts are [35,36,40]:
Fuzzification

â Rule Base—Inference Engine
Fuzzification

â Rule Base—Inference Engine
Defuzzification
 Defuzzification
Figure 20.
Figure Structure of
20. Structure of Fuzzy
Fuzzy Logic
Logic controller.
controller.
During
During the the fuzzification
fuzzificationstep, step,the thecrisp
crispdata,
data, which
which comecome fromfromsomesome measuring
measuring equipment,
equipment, are
are converted into degrees of membership by a lookup in one or several
converted into degrees of membership by a lookup in one or several membership functions. The membership functions. The
fuzzification block matches
fuzzification block matches the the input
input data
data with
with thethe conditions
conditions of of the
the rules
rules to
to determine. There is
determine. There is aa
degree of membership for each linguistic term that applies to the
degree of membership for each linguistic term that applies to the input variable [37]. input variable [37].
The
The rule
rule base
base isis the
the collection
collection of of rules
rules that
that the
the controller
controller should
should meet.
meet. The
The rules
rules are
are in
in “if-then”
“if-then”
format and formally the “If” side is called the conditions and the “Then” side
format and formally the “If” side is called the conditions and the “Then” side is called the conclusion. is called the conclusion.
The
The computer
computer is is able
able to to execute
execute the the rules
rules and
and compute
compute aa control
control signal
signal depending
depending on on the
the measured
measured
inputs
inputs error
error (e)
(e) and
and the the change
change in in error
error d(e)
d(e) [37]. The number
[37]. The number of of rules
rules in
in aa fuzzy
fuzzy logic
logic controller
controller
depends
depends on on the
the number
number of of input
input andand output
output variables
variables and
and the
the number
number of of membership
membership functions.
functions.
During
During the thefuzzification
fuzzification cycle,
cycle,all the
all actions that have
the actions thatbeen
haveactivated are combined
been activated and converted
are combined and
into a single non-fuzzy output signal which is the control signal
converted into a single non-fuzzy output signal which is the control signal of the system. of the system. The output levels
The output
depend on the on
levels depend rules
thethat thethat
rules systems have in place,
the systems have in and the positions
place, depend on
and the positions the non-linearity’s
depend on the non-
existing to existing
linearity’s the systems.to theDeveloping the controlthe
systems. Developing curve of the
control system,
curve representing
of the the I/O relation
system, representing of
the I/O
the systems, and based on the information, the output degree of the membership
relation of the systems, and based on the information, the output degree of the membership function function can be
defined in orderintoorder
can be defined minimize the effectthe
to minimize of the non-linearity
effect [37]. The [37].
of the non-linearity most The
usual membership
most functions,
usual membership
for variable speed error and control output, are the triangular and the basic considerations of the
control philosophy are [37]:
â If both error and change of error is zero, then maintain the present control setting.
â If the error is not zero but is approaching this value at a satisfactory rate, then maintain the
present control setting.
â If the error is growing then change the control signal output depending on the magnitude and
sign of the error, to force the error towards zero.
The speed control of a separately excited DC motor can be implemented using the armature
voltage variation, the field flux variation or the armature resistance variation [40]. The armature
resistance speed control is not accepted because of the high-power losses.
In Reference [38], the armature voltage and field flux control methods are applied independently
and finally are combined together on a single separately excited DC motor. These methods use also the
fuzzy technique. The speed below the rated speed was achieved using the armature voltage speed
control technique by varying the armature voltage in the constant torque region. The speed above the
rated speed was achieved using the field current speed control technique by varying the field voltage
in the constant power region. In the constant power region, field flux should be reduced to achieve
speed above the rated speed [41]. According to the simulation results only the combined armature
voltage and field control method is accurate and accepted. However, the combined armature voltage
and field current control method is capable of an accurate description of the combined effects of the
armature and field excitation to the motor. Also, the proposed combined technique offers the unique
Electronics 2020, 9, 887 20 of 26
characteristic of controlling motor’s speed below and above the rated speed. Additionally, it was
observed that the armature voltage control method is closer to the proposed combined method with
respect to the delay time, rise time and settling time, whereas the field control technique presents the
highest overshoot in all cases and for different reference speed values.
3.2.2. Artificial Neural Network Controllers (ANNC)

Artificial Neural Network (ANN) is a form of artificial intelligence for emulating the human
thinking process. It is different from the Fuzzy Logic Techniques which are rule-based techniques.
There are a lot of different ANN controllers with the three most popular being:
â Model predictive control

â NARMA-L2 control
â Model Reference control
ANN applications, on electrical drive systems, have got as basic idea to identify the mathematical
model of the system’s motor and using this model finally to control motor’s speed [42]. In some cases,
the adaptive control technique is applied in which the inverting forward ANN is used with some
input parameters [43]. ANN techniques can be used in combination with PID control techniques,
as proposed in [26], where two control strategies are proposed.
Neural networks permit the development of quantitative expressions without compromising the
known complexity of the problem. They resemble the human brain in the following two ways [44]:
â A neural network acquires knowledge through learning.

â A neural network’s knowledge is stored within inter-neuron connection strengths known as
synaptic weights.
The reasons for using the ANN controllers broadly are the following [45]:
â All the ANN signals are transmitted in one direction, the same as in automatically control system.
â The ability of ANNs to learn the sample.
â The ability to create the parallel signals in analog as well as in the discrete system.
â The adaptive ability.
The structure of a neural network is composed of a large number of interconnected processing

elements, named neurons, which work in unison to solve specific problems. ANNs learn by example
like people do. Each ANN structure can be applied to solve a specific problem and is configured
through a learning process [44,45]. Neural networks have got the ability of being applied both on
linear and non-linear systems, and they can extract system’s relationships directly from the data
being modeled.
Generally, a neural control scheme consists of two parts. The first part is the neural estimator
which is used to estimate the motor speed. The second part is the neural controller which is used to
generate a control signal for a converter. These two parts must be trained, in order to work as better
as they can. This is the reason that these two parts are passing through a training procedure. In [45],
this training procedure is applied using the Levenberg-Marquardt back propagation algorithm. To
train ANNs, the input and output datasets have to be determined first, and then design the ANN’s net
by optimizing the number of hidden layers, the number of neurals of each layer as well as the input
and output number and the transfer function.
A neural network consists of junctions which are connected with LINKS, also called processing
units. For each junction, a number is attached called the weight of junction. The weights are the tools
for the long-distance information storing in the neutral network. The learning process occurs with the
appropriate modification of weights. These weights are modified so that the network input/output
behavior is in consonance with the environment which provides the data. A single input neuron
consists of a scalar input ‘p’ multiplied by the scalar weight ‘w’ to calculate the product ‘w × p’ which
input and output number and the transfer function.
A neural network consists of junctions which are connected with LINKS, also called processing
units. For each junction, a number is attached called the weight of junction. The weights are the tools
for the long-distance information storing in the neutral network. The learning process occurs with
the appropriate
Electronics 2020, 9, 887 modification of weights. These weights are modified so that the network 21 of 26
input/output behavior is in consonance with the environment which provides the data. A single input
neuron consists of a scalar input ‘p’ multiplied by the scalar weight ‘w’ to calculate the product ‘w ×
is
p’fed to the
which summer
is fed to the along
summer with biaswith
along ‘b’ multiplied by ‘1’ [44].
bias ‘b’ multiplied by Figure
‘1’ [44].20 illustrates
Figure the basicthe
20 illustrates neural
basic
network diagram.
neural network diagram.
In
In the
the basic ANN diagram
basic ANN diagram showed
showedin inFigure
Figure21,
21,the
thenetnetinput
input is is
‘w‘w× p×+pb’+andb’ and the output
the output is ‘a
is ‘a = f(w × p + b)’. The ‘f’ is the transfer function while the ‘w’ and ‘b’
= f(w × p + b)’. The ‘f’ is the transfer function while the ‘w’ and ‘b’ are adjusted through learning are adjusted through
rule
learning
[44]. In [44]rule [44]. In [44]and
simulation simulation
comparison and comparison
between PID, between
PI andPID, ANN PI controllers
and ANN controllers are
are presented.
presented.
According According to theANNC
to the authors, authors,haveANNC
betterhave better performance
performance by reducing by the
reducing
steadythe steady
state state
error, the
error, the maximum overshoot the rise time and the settling time. Also,
maximum overshoot the rise time and the settling time. Also, the ANNC have more sensitive the ANNC have more sensitive
responses
responsesagainst
againstload
loaddisturbances
disturbancesin incomparison
comparisonto tothe
theclassical
classicalPI PIand
andPIDPIDcontrollers.
controllers. The
The higher
higher
rise
rise time of the system, using ANN controller, compared to the rise times of the systems using
time of the system, using ANN controller, compared to the rise times of the systems using the
the
other
other controllers,
controllers,isisconsidered
consideredaadisadvantage.
disadvantage.
Figure21.
Figure 21. Basic
Basic Neural
Neural Network
Network diagram
diagram [44].
[44].
Figure
Figure 22 22 illustrates
illustrates thethe basic
basic structure
structure of of an
anANN ANNmodel.model. Such a model model is is consisted
consisted of of at
at
minimum
minimumthree threelayers.
layers.TheThefirst
firstlayer
layerisisthe
theinput
inputlayer.
layer.At Atthis
thislayer,
layer,allallthe
themeasured
measured parameters
parameters of of
a
motor
a motor control
controlsystem
system areare
fedfed
intointo
the the
ANN ANNstructure.
structure.TheThe
inputs inputspasspass
through the hidden
through layers
the hidden and
layers
the
andoutput layerlayer
the output is finally produced
is finally in order
produced to send
in order control
to send signalsignal
control to thetopower electronics.
the power In this
electronics. In
specific structure
this specific (Figure
structure 22), a 22),
(Figure three-layer scheme
a three-layer is used.
scheme is The
used. inputs are theare
The inputs estimated speed ω
the estimated p (k),
speed
the terminal
ωp(k), voltage voltage
the terminal Vt (k), the armature
Vt(k), current Icurrent
the armature a (t) andIthe
a(t) reference speed ωrspeed
and the reference (t) as they
ωr(t)are
as selected
they are
by the scheme
selected by thedesigner in [45]. Figure
scheme designer in [45].23Figure
illustrates the ANN-
23 illustrates thetraining
ANN- structure used by the
training structure usedauthors
by the
of [45] inof
authors order
[45] into order
calculate the estimated
to calculate speed ωspeed
the estimated p (k). For Figures
ωp(k). 22 and 22
For Figures 23 and
f1 and f2 and
23 f1 are the input
f2 are the
and hidden layers’ tansig-sigmoid activation
input and hidden layers’ tansig-sigmoid activation functions, functions, V t and IV are the measured terminal
a t and Ia are the measured terminal voltage
and armature
voltage
Electronics and 8,current
2019, armature
x FOR PEERand Alpha
current
REVIEW and output
Alphaisoutput
the control
is thesignal
control forsignal
the controller.
for the controller. 22 of 27
Figure
Figure 22. Basic ANN
22. Basic ANN structure
structure consisted
consisted of
of 33 layers
layers [45].
[45].
Moreover, Table 3 presents the main characteristics for the control techniques applied in
DC motor drive systems, which are described in this paper, and which namely are the classic
PID(Proportional-Integral-Derivative) controllers and the intelligent controllers. About the intelligent
controllers, these papers presented the most known categories, which are the fuzzy logic controllers
and the controllers, based on artificial neural networks.
Figure 23. Basic ANN-training structure consisted of 3 layers [45].
Moreover, Table 3 presents the main characteristics for the control techniques applied in DC
motor drive systems, which are described in this paper, and which namely are the classic
Electronics 2020, 9, 887 22 of 26
Figure 22. Basic ANN structure consisted of 3 layers [45].
Figure
Figure 23.
23. Basic
Basic ANN-training
ANN-training structure
structure consisted
consisted of
of 33 layers
layers [45].
[45].
Table 3. Control techniques for brushed DC motors drives.

Moreover, Table 3 presents the main characteristics for the control techniques applied in DC
motor drive systems, which are described in this Highpaper, and which
performance namely
motor drive are
system thehave:
must classic
Control objectives:
PID(Proportional-Integral-Derivative) controllers •andFast thespeed
intelligent controllers.
transient response About the
• Closed-loop stability
intelligent controllers, these papers presented the most known
• Fast loadcategories, which are the fuzzy logic
transient response
• Adequate performance
controllers and the controllers, based on artificial neural
• networks.
High efficiency
• Adequate robustness
• Effective control
Table 3. Control techniques for•brushed
HighDCreliability
motors drives.
Classic PID Controllers
High performance motor drive system
• Control
Modeling of individual parts of the system must have:
objectives: • PID calculation and optimization using:
• Requirement for a relatively accurate model
• Fast speed transient response
• • Closed-loop
Response and stability
stability control applying Nyquist - Ziegler-Nichols method
• Fast load transient response
• Adequate performance
criterion and/or Bode diagrams - Cohen-Coon method
• • Adequate
Well knownrobustness
procedure • High
- efficiency
Manual tuning method
• Effective
- control
Software tools (MATLAB tools)
• High
Intelligent Controllers reliability
Fuzzy Logic
Classic PID Controllers
Artificial Neural Networks (ANN)
• Modeling of individual parts of the system • PID calculation and optimization using:
• • Requirement
Implements anfor if-then programming
a relatively formmodel
accurate for • Have got as basic idea to identify the
the various working conditions to control the Ziegler-Nichols
-mathematical method
model of the system’s motor
• Response
motor speed
and stability control applying - Cohen-Coon method
• Adaptive control technique is applied
• Nyquist criterion
High precision and/or
modeling is Bode diagrams
not required • -AllManual
the ANNtuning
signalsmethod
are transmitted in one
• • Well known
Systematic wayprocedure
to take under consideration -direction,
Softwarethetools
same (MATLAB
as in automatically
tools)
model’s unintentional or/and control system
Intelligent Controllers
intentional uncertainty • Ability of ANNs to learn the sample
• Fuzzy Logic
Inherent approximation capability Artificial Neural Networks (ANN)
• Minimum overshoot of the speed output
• • Implements
High degree of antolerance
if-then programming form • • Have
Good got
speedasresponse
basic idea
forto identify
constant thetorque
load
• for the various
Smooth operationworking conditions to mathematical
which finally canmodel
reject of the system’s
motor’s perturbation
• control
Learningthe motor speed
ability • motor
Simplified versions for the independent control
• High precision modeling is not required • Adaptive
of flux andcontrol
torque using cascade
technique is structures
applied
• Systematic way to take under consideration • All the ANN signals are transmitted in
4. Semi-Quantitative Comparative
model’s unintentional Assessment
or/and intentional one direction, the same as in
uncertainty
In automatically
this section a semi-quantitative comparative assessment forcontrol
brushedsystem
DC motor drives is
• Inherent approximation capability •
attempted. As it has been presented in the previous sections, Tables 1 and 2 the
Ability of ANNs to learn samplethe basic
presents
characteristics of the controlled rectifier-fed and chopper-fed motor drives suitable for brushed DC
motor drives.
Table 4 is an effort to summarize, in a compact shape, the advantages and disadvantages of
different converter topologies combined with control methods, suitable for driving brushed DC motors.
This assessment takes into consideration some of the most important or most commonly referring
characteristics like quadrants of operation, power level, operation frequency, energy efficiency, output
ripple current, transient response and level of complexity.
Electronics 2020, 9, 887 23 of 26
Table 4. Comparative assessment of converters for brushed DC motor drives.
Parameters Quadrants of Ripple Transient

Power Level Efficiency Frequency Complexity
Converter Operation Current Response
1Φ half wave 1 low medium to high low high slow low
1Φ asymmetrical converter 1 medium medium to high low high slow low
1Φ fully-controlled converter 2 medium medium to high low high slow low
1Φ fully-controlled dual converter 4 medium medium to high low high slow low
3Φ half-controlled converter 1 Medium to high medium to high low high slow low to medium
3Φ fully-controlled converter 2 high to very high medium to high low high slow low to medium
3Φ fully-controlled dual converter 4 high to very high medium to high low high slow low to medium
1-switch chopper 1 low to medium medium medium low fast low to medium
2Q chopper 2 low to medium medium medium low fast low to medium
H-Bridge chopper 4 medium medium medium low fast low to medium
2Q-ZVMR converter 2 low to medium high high very low very fast medium to high
2Q ZVT converter 2 low to medium high high very low very fast medium to high
2Q ZCT converter 2 low to medium high high very low very fast medium to high
4Q ZVT converter 4 medium high high very low very fast high
4Q ZCT converter 4 medium high high very low very fast high
Electronics 2020, 9, 887 24 of 26
As the quantitative approach on such effort is not possible as the values of the above mentioned
characteristics depend on the size and type of the motor, the application and other factors, this approach
is based on a qualitative and semi-quantitative representation of these important and critical operational
parameters, aiming to deliver a decision support view in the scientific and engineering community.
5. Conclusions
This paper offers an extensive review of the basic topologies, and control methods and techniques
applied in conventional and modern brushed DC motor drives, for industrial and automobile
applications. The field of applications of each power circuit topology, include the advantages,
disadvantages, as well the control techniques presented for each case.
DC motor drives are still used and have evolved due to the exceptional characteristics of DC
motors. Different and varied application requirements make the use of drive systems mandatory, with
less or more complex control systems. The modern industry uses complex power systems to meet the
growing needs of society for newly developed products, and faster production rates, respectively.
The maturity of advanced control techniques such as Fuzzy Logic (FL) and Artificial Neural
Network (ANN) have made possible to embed them into modern, high standard controllers specialized
for high efficiency motor drives.
Additionally, the new legislation on the eco-design of the systems and the high efficiency
requirements, have led to the development of new strategies in the field of electric motor control.
Author Contributions: Conceptualization, D.A.B., G.C.I. and C.S.P.; methodology, G.C.I. and C.S.P.; validation,
G.V. and S.D.K.; formal analysis, D.A.B., G.C.I. and C.S.P.; investigation, D.A.B.; resources, D.A.B., G.C.I.
and G.A.V.; data curation, D.A.B., G.C.I. and S.D.K.; writing—original draft preparation, D.A.B. and G.C.I.;
writing—review and editing, D.A.B., G.C.I. and C.S.P.; visualization, D.A.B., G.C.I. and C.S.P.; supervision,
G.C.I.; project administration, G.C.I. and C.S.P. All authors have read and agreed to the published version of
the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Chapman, S.J. Electric Machinery Fundamentals; WCB/McGraw-Hill: New York, NY, USA, 1998.
2. Mohan, N.; Undeland, T.M. Power Electronics, Converters, Applications and Design; John Wiley & Sons:
New York, NY, USA, 1995.
3. Manias, S.N. Power Electronics; Simeon Publications: Athens, Greece, 2012. (In Greek)
4. Malatestas, P. Motor Drives; Tziolas Publications: Athens, Greece, 2013. (In Greek)
5. Williams, B.W. Power Electronics: Devices, Drivers, Applications, and Passive Components; University of
Strathclyde: Glasgow, UK, 2006.
6. Nedeljkovic, M.; Stojiljkovic, Z. Fast current control for thyristor rectifiers. IEE Proc. Electr. Power Appl. 2003,
150, 636–638. [CrossRef]
7. Consoli, A.; Cacciato, M.; Testa, A.; Gennaro, F. Single chip integration for motor drive converters with
power factor Capability. IEEE Trans. Power Electron. 2004, 19, 1372–1379. [CrossRef]
8. Daigavane, M.; Suryawanshi, H.; Khan, J.A. Novel Three Phase Series-Parallel Resonant Converter Fed
DC-Drive System. J. Power Electron. 2007, 7, 222–232.
9. Gupta, R.; Lamba, R.; Padhee, S. Thyristor based speed control techniques of DC motor: A comparative
analysis. Int. J. Sci. Res. Publ. 2012, 2, 1–6.
10. Chau, K.T.; Lee, Y.S.; Ioinovici, A. Computer-aided modeling of quasi-resonant converters in the presence of
parasitic losses by using MISSCO concept. IEEE Trans. Ind. Electron. 1991, 38, 454–461. [CrossRef]
11. Chan, C.C.; Chau, K.T. A new zero-voltage-switching dc/dc boost converter. IEEE Trans. Aerosp. Electron.
Syst. 1993, 29, 125–134. [CrossRef]
12. Chau, K.T. New Constant-frequency multi-resonant boost convertor. IEE Electron. Lett. 1994, 30, 101–102.
[CrossRef]
Electronics 2020, 9, 887 25 of 26
13. Chong, C.C.; Chan, C.Y.; Foo, C.F. A quasi-resonant converter-fed dc drive system. In Proceedings of the
European Power Electronics Conference, Brighton, UK, 13–16 September 1993; pp. 372–377.
14. Hua, G.; Lee, F.C. Soft-switching Techniques in PWM Converters. IEEE Trans. Ind. Electron. 1995, 42, 595–603.
[CrossRef]
15. Luo, F.L.; Jin, L. Two-quadrant DC/DC Soft-switching Converter. Proc. IEEE Power Electron. Spec. Conf. 2000,
1, 173–178.
16. Uma, G.; Chellamuthu, C. Modeling and Design of Fuzzy Speed Controller for Constant Frequency Zero
Current Switched Converter Fed DC Servo Motor for Battery Operated Vehicles. Proc. Int. Conf. Power Syst.
Technol. 2000, 1, 211–215.
17. Ching, T.W. Soft-switching converters for electric vehicle propulsion. J. Asian Electr. Veh. 2007, 5, 1019–1026.
[CrossRef]
18. Chau, K.T.; Ching, T.W.; Chan, C.C. Constant frequency multi-resonant converter-fed dc motor drives. In
Proceedings of the IEEE Industrial Electronics, Control, and Instrumentation Conference, Taipei, Taiwan, 9
August 1996; pp. 78–83.
19. Chau, K.T.; Ching, T.W. A new two-quadrant zero voltage transition converter for dc motor drives. Int. J.
Electron. 1999, 86, 217–231. [CrossRef]
20. Ching, T.W.; Chau, K.T. A new two-quadrant zero current transition converter for dc motor drives. Int. J.
Electron. 2001, 88, 719–735. [CrossRef]
21. Ching, T.W. Four-quadrant zero-voltage-transition converter-fed dc motor drives for electric propulsion.
J. Asian Electr. Veh. 2005, 3, 651–656. [CrossRef]
22. Ching, T.W. Four-quadrant zero-current-transition converter-fed dc motor drives for Electric Propulsion.
J. Asian Electr. Veh. 2006, 4, 911–918. [CrossRef]
23. Grassi, E.; Tsakalis, K. PID Controller Tuning by Frequency Loop-Shaping: Application to Diffusion Furnace
Temperature Control. Proc. IEEE Trans. Control Syst. Technol. 2000, 8, 842–847. [CrossRef]
24. Loucif, F. DC Motor Speed Control using PID Controller. In Proceedings of the International Conference on
Control, Automation and Systems (ICCAS), KINTEX, Gyeonggi-Do, Korea, 2–5 June 2005.
25. Ioannidis, G.C.; Kaminaris, S.D.; Psomopoulos, C.S.; Tsiolis, S.; Pachos, P.; Villiotis, I.; Malatestas, P. DC
Motor Drive Applying Conventional and Fuzzy Based PI Control Techniques. JARR 2015, XV, 1–10.
26. Singh, B.; Prakash, S.; Pandey, A.S.; Sinha, S.K. Intelligent PI Controller for Speed Control of D.C. Motor.
Int. J. Electron. Eng. Res. 2010, 2, 87–100.
27. Mishra, G.K.; Pandey, A.K.; Maurya, A. Combined Armature and Field Speed Control of DC Motor for
Efficiency Enhancement. SSRG Int. J. Electr. Electron. Eng. (SSRG-IJEEE) 2014, 1, 34–39.
28. Zaccarian, L. DC Motors: Dynamic Model and Control Techniques; Lecture Notes. Roma, Italy, 2012. Available
online: http://control.disp.uniroma2.it/zack/LabRob/DCmotors.pdf (accessed on 15 May 2020).
29. Haishui, Z.; Dahu, W.; Tong, Z.; Keming, H. Design on a dc motor speed control. In Proceedings of the
International Conference on Intelligent Computation Technology and Automation, Changsha, China, 11–12
May 2010. [CrossRef]
30. Aruna, K.; Pravin, M.R.; Ramachandran, J.; Sudheer, S.; Sathiyasekar, K. Design and Implementation of PI
Controller for the Speed Control of DC Motor in Cooling System. Int. J. Eng. Trends Technol. (IJETT) 2014.
[CrossRef]
31. Castano, L.F.; Osorio, G.A. Design of a FPGA based position PI servo controller for a DC Motor with dry
friction. In Proceedings of the 2011 VII Southern Conference on Programmable Logic (SPL), Cordoba,
Argentina, 13–15 April 2011; ISBN 978-1-4244-8848.
32. Zhang, D.; Li, H.; Collins, E.G. Digital Anti-Windup PI Controllers for Variable-Speed Motor Drives Using
FPGA and Stochastic Theory. Proc. IEEE Trans. Power Electron. 2006, 21, 1496–1501. [CrossRef]
33. Zenginobuz, G.; Bilgin, H.F.; Kose, K.N.; Ermis, M. A Unity Power Factor Buck-Type PWM Rectifier for
Medium/High- Power DC Motor Drive Applications. Proc. IEEE Trans. Ind. Appl. 2002, 38, 1412–1425.
34. Chen, J.; Tang, P.C. A Sliding Mode Current Control Scheme for PWM Brushless DC Motor Drives. Proc. IEEE
Trans. Power Electron. 1999, 14, 541–551. [CrossRef]
35. Lim, C.S.; Levi, E.; Rahim, N.A.; Hew, W.P. A Comparative Study of Synchronous Current Control Schemes
Based on FCS-MPC and PIPWM for a Two-Motor Three-Phase Drive. IEEE Trans. Ind. Electron. 2014, 61,
3867–3878. [CrossRef]
Electronics 2020, 9, 887 26 of 26
36. Sen, P.C.; Namuduri, C.S.; Nandam, P.K. Evolution of control techniques for industrial drives. In Proceedings
of the International Conference on Power Electronics, Drives and Energy Systems for Industrial Growth,
New Delhi, India, 8–11 January 1996. [CrossRef]
37. Rai, J.N.; Singhal, M.; Nandwani, M. Speed Control of Dc Motor Using Fuzzy Logic Technique. IOSR J. Electr.
Electron. Eng. (IOSR-JEEE) 2012, 3, 41–48. [CrossRef]
38. Sadiq, A.A.; Bakare, G.A.; Anene, E.C.; Mamman, H.B. A Fuzzy-Based Speed Control of DC Motor Using
Combined Armature Voltage and Field Current. In Proceedings of the 3rd IFAC International Conference on
Intelligent Control and Automation Science 2013, Chengdu, China, 2–4 September 2013.
39. Zadeh, L.A. Fuzzy Sets, Fuzzy Logic, Fuzzy Systems; World Scientific Press: Singapore, 1996; ISBN 9810224214.
40. Mendel, J. Fuzzy logic systems for Engineering applications. Proc. IEEE 1995, 83, 345–377. [CrossRef]
41. Mustapha, A. Speed Control of Dc Motor using Combined Armature and Field Control. Unpublished
Ph.D. Thesis, Brno University of Technology, Brno, Czech Republic, 2005.
42. Dzung, P.Q.; Phuong, L.M. ANN- Control system dc motor. IEEE Trans. Neural Netw. 2002, 82–90.
Available online: http://www.nsl.hcmus.edu.vn/greenstone/collect/hnkhbk/import/Dien_tu/3_SectionA/10_
Phan%20Quoc%20Dzung,%20Le%20Minh%20Phuong_ANN%20-%20CONTROL%20SYSTEM%20DC%
20MOTOR.pdf (accessed on 18 May 2020).
43. Weerasoory, S.; AI-Sharkawi, M.A. Identification and control of d.c. motor using back propagation neural
networks. IEEE Trans. Energy Convers. 1991, 6, 663–669. [CrossRef]
44. Kumar, S.B.; Ali, M.H.; Sinha, A. Design and Simulation of Speed Control of DC Motor by Artificial Neural
Network Technique. Int. J. Sci. Res. Publ. 2014, 4, 1–4.
45. Madhusudhana Rao, G.; Sanker Ram, B.V. A Neural Network Based Speed Control for DC Motor. Int. J.
Recent Trends Eng. 2009, 2, 121–124.
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Electronics 09 00887 v2 PDF

Uploaded by

Document Information

Original Title

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

Electronics 09 00887 v2 PDF

Uploaded by

Copyright:

Available Formats

electronics

Electronics 2020, 9, 887; doi:10.3390/electronics9060887 www.mdpi.com/journal/electronics

the power source.

â Classical PID Controllers

braking mode or reverse regeneration mode.

eebb rrff ,, LLff Vf

TT1111 TT1313 TT1515 RRαα TT2222 TT2626 TT2424

Table 1. Main characteristics of controlled rectifier-fed drives.

• Explicit ripple in • Low • 0.5 kW to

2.2. Chopper—Fed DC Motor Drives

2.2.1. Hard-Switching Converters for DC Drives

I Safe Operating Area

Figure 10. Hard and soft switching characteristics.

where VVoo,, eebb ,, V

Vo = (2D − 1)Vin , (13)

and for D ≥ 0.5

2.2.2. Soft-Switching Converters for DC Drives

Figure 13 illustrates a two-quadrant (2Q) Zero-Voltage Multi-Resonant (ZVMR) converter [18].

Figure 13. Zero-voltage multi-resonant converter fed DC motor drive.

Figure 15. Zero-current-transition converter fed DC motor drive.

Figure 16. 4Q-ZVT converter fed DC motor

Figure 17. 4Q-ZCT converter fed DC motor drive.

2.3. Comparative Evaluation of DC Motor Drives

Table 2. Main characteristics of chopper-fed drives.

• High order harmonics • 0.5 kW to

• Typical switching frequency • Considerable transient effects

• Usage of LC resonant circuits

3. Control Techniques for DC Motor Drives

3.1. Classic PID Controllers

â Proportional (P) based control technique

3.2. Intelligent Controllers

3.2.1. Fuzzy Logic Controllers

3.2.2. Artificial Neural Network Controllers (ANNC)

â Model predictive control

â A neural network acquires knowledge through learning.

The structure of a neural network is composed of a large number of interconnected processing

Figure 23. Basic ANN-training structure consisted of 3 layers [45].

Table 3. Control techniques for brushed DC motors drives.

Table 4. Comparative assessment of converters for brushed DC motor drives.

Parameters Quadrants of Ripple Transient

You might also like