Professional Documents
Culture Documents
CHAPTER 1
INTRODUCTION
1.1 GENERAL
literature review and a new error integrative algorithm suitable for high
frequency switched three-phase-to-three-phase matrix converter are presented
along with the objective and organization of the thesis.
AC-to-AC
Converter topologies
Va Vb Vc
Equation (1.2) gives the constraints namely that the inputs are not
short-circuited and the outputs are not open-circuited
d Aj + d Bj +d Cj =1 (1.2)
where, dij is the duty-cycle of the switch Sij. Equation (1.2) being less than one
indicates an open-circuit of the current source and Equation (1.2) being
greater than one indicates a short-circuit of the voltage source.
V out = T × V in
V a S Aa SAb SAc VA
[ ][
V b = S Ba SBb SBc × VB
Vc SCa SCb SCc VC ][ ] (1.3)
where,
Va S Aa SAb SAc VA
[] [ ]
Vout = Vb , T= SBa SBb SBc and V in = VB
Vc SCa SCb SCc VC []
6
I in = T T × I out
where,
IA SAa S Ba SCa Ia
[] [ T
] []
I in = I B , T = SAb S Bb SCb and I out = I b .
IC SAc S Bc SCc Ic
VA, VB, VC and Va, Vb, Vc are the input and the output phase voltages
respectively and IA, IB, IC and Ia, Ib, Ic are the input and the output currents
respectively.
LA S1+ S1- IL
VA
CA
RL - Load
VB LB S2+ S2-
CB
BS S2
Step 2 : Then, the switch in the incoming BS that would conduct the
current in the same direction is turned ON. This is done to form
a path for the load current to continue flowing either at the point
when the next switch of the incoming BS is gated ON or when
the conducting switch of the outgoing BS is turned OFF.
(i) Switch S1- is turned OFF, (ii) Switch S2+ is turned ON,
(iii) Switch S1+ is turned OFF, and (iv) Switch S2- is turned ON.
1 1 0
0 0 0
iL > 0 0 1 1
0 0 0
S1+ S1+ 0
1 0
1 S1- S1-
S2+ S2+ 1
0 1
0 S2- S2-
0 0 0
1 1 0
iL < 0
0 0 0
0 1 1
Since the commonly used Hall effect sensors are prone to produce
uncertain results in high power and low current applications, it is difficult to
reliably determine the direction of current for commutation. To avoid this
problem, a technique named as voltage commutation that uses the voltage
across the bidirectional switch for measurement of the direction of current
(Wheeler et al 2002b) has been developed. Later, a technique utilizing the
zero vectors (Mahlein et al 2002) to avoid commutation error, when the line
voltage is zero, was proposed for robust commutation. However, the input
current in this technique was found distorted compared to the voltage and
current commutation techniques because of utilizing a different switching
sequence.
MIN-MID-MAX modulation
VA
SAp SAn
VB
SBp SBn
VC
SCp SCn
P N LLa RLa
Sap San
LLb RLb
Vb
Sbp Sbn
LLc RLc
Vc
Scp Scn
The ISVM method obtains the required output voltages through two
analytically independent stages or transformations of the input voltages. This
requires representing the switching function T, given by Equation (1.3), as the
product of the rectifier switching function and the inverter switching function,
as given in Equation (1.5). In the first stage, the three-phase AC voltage is
converted to an average DC bus voltage of constant value. In the second
stage, this constant average DC bus voltage is converted to a three-phase AC
voltage of the required frequency and amplitude. Equations (1.6) and (1.7)
give the equations for the output voltages, output currents, input voltages, and
input currents with respect to the fictitious DC bus
T= T I × TR
[ ][ ][ S S S
SBa SBb SBc = S bp Sbn × Ap Bp Cp
SCa SCb SCc Scp Scn
SAn SBn SCn ] (1.5)
where,
[ ] [ ] S S
[ S
T= SBa SBb SBc , T I = Sbp Sbn and TR = Ap Bp Cp
SCa SCb SCc Scp Scn
SAn SBn SCn ]
V a Sap San
[ ][ ][ ]
Vc Scp Scn
V
V b = S bp Sbn × DC+
V DC-
(1.6)
Ia
[ ][
I DC-
=
][]
I DC+ Sap Sbp Scp
San Sbn Scn
× Ib
Ic
I A SAp S An
[ ][ ][ ]
I C SCp SCn
I
I B = SBp SBn × DC+
I DC-
(1.7)
15
VA
V DC+ SAp SBp SCp
[ ][
V DC-
=
SAn SBn SCn ][
× VB
VC ]
The restrictions on the switch states that were analyzed in section
1.3 can now be applied to the rectifier and the inverter sections independently.
The switches of the inverter stage, associated with lines P and N in Figure 1.7,
cannot be simultaneously closed but the switches of the converter stage,
associated with lines P and N, can be closed. Equations (1.8) and (1.9)
express these restrictions mathematically.
P DC = P IN
3 I
VDC = VIN IN cos(φ in )
2 I DC
3
V DC = V IN m c cos( φin ) (1.10)
2
16
P DC = P OUT
3 V OUT
I DC = IOUT cos(φ out )
2 V DC
3
I DC = IOUT m v cos( φout ) (1.11)
2
Equations (1.12) and (1.13) express the inputs IIN, VIN and the
outputs VOUT, IOUT as space vectors.
2π 4π 2π 4π
2 2
( j j
VIN = V A + VB e 3 + VC e 3 ,
3
) ( j j
I IN = I A + IB e 3 + IC e 3
3
)
(1.12)
2π 4π 2π 4π
2 2
VOUT =
3
( j j
Va + V b e 3 + V c e 3 , ) I OUT =
3
( j j
Ia + I b e 3 + I c e 3 )
(1.13)
Tables 1.1 and 1.2 present the space vectors and the relevant
switching states of the converter and the inverter stages.
Table 1.1 Switching states and input currents for the converter stage
m SCm IA IB IC | I IN | ∠ I IN
2 π
1 [1 -1 0] I DC - I DC 0 I DC -
√3 6
2 π
2 [1 0 -1] I DC 0 - I DC I DC +
√3 6
2 π
3 [0 1 -1] 0 I DC - I DC I DC +
√3 2
17
2 5π
4 [-1 1 0] - I DC I DC 0 I DC +
√3 6
2 5π
5 [-1 0 1] - I DC 0 I DC I DC -
√3 6
2 π
6 [0 -1 1] 0 - I DC I DC I DC -
√3 2
7 [(1, -1) 0 0] 0 0 0 0 -
8 [0 (1, -1) 0] 0 0 0 0 -
9 [0 0 (1, -1)] 0 0 0 0 -
Table 1.2 Switching states and output voltage for the inverter stage
Note: 1 in the SCm represents the upper switch in an arm being ON,
-1 in the SCm represents the lower switch in an arm being ON, (1,-1) in the SCm
represents both the switches in an arm being ON and 0 in the S Cm represents
both the switches in an arm being OFF.
synthesized as a vector sum of any two adjacent active vectors (I 1 to I6) and
one of the zero vectors (I0a, I0b, I0c).
Vb Vb
V2 [110] IM ejπ/2
V3 [010] Zero Vectors
I3 [01-1]
VM ejπ/3 V0 [000]
VM ej2π/3
1 V7 [111] IM ejπ/6
2 1
I2 [10-1]
2 0 V1 [100] IM ej5π/6
V4 [011] I4 [-110]
Vβdvβ VM e-j0
VM ejπ 3
Va Iβ dIβ 0 Va
3 θv VM
Vαdvα
θc IM
Iα dIα
5 5
IM e-j5π/6 4
IM e-jπ/6
4 I5 [-101]
V5 [001] Zero Vectors I1 [1-10]
V6 [101]
VM e-j2π/3 VM e-jπ/3
I0a [(1,-1)00]
Vc Vc IM e-jπ/2
I0b [0(1-1)0] I6 [0-11]
Figure 1.8 (a) Inverter voltage hexagon and (b) rectifier current hexagon
Vy
dyVy
AV
θAv
dxVx Vx
Tx π
dx = = m i sin( - θAv ) (1.14)
Ts 3
Ty
dy = = m i sin ( θAv ) (1.15)
Ts
Tz
dz = = (1 - dx - dy ) (1.16)
Ts
where, θAV represents the angle of the reference arbitrary vector within that
sector, mi is the modulation index defining the magnitude of the arbitrary
vector AV.
T vα π
d vα = = m v sin( - θ v )
Ts 3
Tvβ
d vβ = = m v sin ( θv )
Ts
T v0
d v0 = = (1 - d vα - d vβ ) (1.17)
Ts
Tcα π
d cα = = m c sin( - θc )
Ts 3
20
Tcβ
d cβ = = m c sin ( θc )
Ts
Tc0
d c0 = = (1 - dcα - d c0 ) (1.18)
Ts
The combination of the six active states of the rectifier with the six
active states of the inverter produces a set of 36 active states. These states can
be grouped into 18 pairs of equivalent states. The states in each pair are
equivalent because both connect the same input to the same output. For
example, the combination of the rectifier state m=1 and inverter state m=6
given in Tables 1.1 and 1.2 is equivalent to the combination of the rectifier
state m=4 and inverter state m=3. Both result in the connection of the output
lines a and c to the input line A and the output line b to the input line B. Both
these combinations are the same state (ABA) in matrix converter, as shown in
Figure 1.10.
A B C A B C
a b c
(c)
21
Equations (1.19) to (1.23) give the duty cycles for the matrix
converter, which are derived from the product of the inverter duty cycles and
the rectifier duty cycles (Cha 2004).
π π T vα cα
d vα cα =d vα × d cα = m v sin ( 3 ) (
- θv × m c sin - θc =
3 Ts) (1.19)
Tvα cβ
d vα cβ = d vα × dcβ = m v sin (3π - θ ) × m sin (π3 - θ ) = T
v c c
s
(1.20)
T vβ cα
d vβ cα = d vβ × d cα = m v sin ( π3 - θ ) × m sin (π3 - θ ) = T
v c c
s
(1.21)
T vβ cβ
d vβ cβ = d vβ ×d cβ = m v sin (3π - θ )× m sin (π3 - θ ) = T
v c c
s
(1.22)
T0
d 0 =1 - d vα cα - d vα cα - d vβ cα - d vβ cβ = (1.23)
Ts
The method minimizes the error between the actual output and the
expected output in a switching period by selecting the switching vector that
provides the minimum square error. In the next switching instant, the
previously calculated minimum square error is added to the present square
errors of the available space vectors, which becomes the input to the
controller that finds the appropriate minimum error switching states. The
23
Ie mj ( t ) = ( I rj ( t ) - I mj ( t ) ) +Iepj ( t ) (1.27)
VA
VDC =
SAp SBp SCp
[
- SAn - SBn - SCn
× VB
VC
][ ] (1.32)
where, Vmj and Imj are respectively the output voltage and the input current
corresponding to the switching matrixes S Im and SCm. Vrj and Irj denote
respectively the reference output voltage and the reference input current; Ve mj
and Iemj are respectively the output voltage error and the input current error
corresponding to switching states SIm and SCm; Vepj and Iepj are respectively
the output voltage error and the input current error due to the previous
switching state and SIon and SCon are respectively the switching matrices used
in the present switching states for the OC and the IC.
assumption does not hold good. This is because, with the application of the
zero vectors of the OC, the IDC becomes zero due to the isolation of the source
from the load. There is a need to recalculate the current error, which could be
used for calculation in the next sampling time. Hence, it is proposed, for the
improvement of MESS technique, to ignore the calculated current error, Ie mj(t-
1) and use Iemj(t-2), when a zero vector is applied on the OC, as given by
Equation (1.33). Figure 1.13 shows the block diagram representation of the
improved MESS technique.
Calculate current error Iem for all Calculate voltage error Vem for all
current vectors SCm voltage vectors SIm
Delay Ts Delay Ts
10
0
0 200 400 600 800 1000
Frequency (Hz)
(a)
10
0
0 200 400 600 800 1000
Frequency (Hz)
(b)
Figure 1.14 Current harmonics (a) MESS technique and (b) improved
MESS technique
200
100
Vab(V)
-100
-200
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Time(s)
(a)
200
100
Vab(V)
-100
-200
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Time(s)
(b)
Figure 1.15 Simulation results of the output phase voltage (a) ISVM
technique and (b) MESS technique
Figure 1.15 shows the MESS switching states that are compared
with the ISVM switching states. It can be observed in Figure 1.16 that in the
ISVM method multiple switchings occur within the sampling frequency when
28
Figure 1.16 Total losses in the CMC with the ISVM and the MESS
techniques
100
50
VABC (V)
0
-50
-100
0 0.02 0.04 0.06 0.08 0.1 0.12
Time(s)
(a)
200
100
Va ( V)
-100
-200
0 0.02 0.04 0.06 0.08 0.1 0.12
Time (s)
(b)
200
100
Vab(V)
-100
-200
0 0.02 0.04 0.06 0.08 0.1 0.12
Time(s)
30
(c)
2
Iabc (A)
-4
0 0.02 0.04 0.06 0.08 0.1 0.12
Time(s)
(d)
2
IABC(A)
-2
-4
0 0.02 0.04 0.06 0.08 0.1 0.12
Time(s)
(e)
Figure 1.17 Simulation results (a) input phase voltage, (b) output phase
voltage, (c) output line voltage, (d) output current and (e)
input current
31
(a)
(b)
(c)
(d)
(e)
Figure 1.18 Hardware results (a) input phase voltage, (b) output phase
voltage, (c) output line voltage, (d) output current and (e)
input current
papers describe the detailed effects of over modulation operation of the matrix
converter (Thuta 2007). The paper discusses four ways of operating the
matrix converter under over modulation (i) output side over modulation,
(ii) input side over modulation with power factor control, (iii) input side over
modulation without power factor control and (iv) simultaneous output and
input side over modulation. Using the over modulation technique, the
theoretical voltage limit of the converter can be increased to 105 % of the
input voltage. It has been proved in Mahlein et al (1999) that some lower
order harmonics are generated at the output voltage and the input current by
the over modulation operation (Wiechmann et al 1997). Over modulation
operation of the matrix converter might cause the resonance of the line side
filter. This might damage the converter if not controlled properly. Thus, it
was concluded in Wiechmann et al (2002) that it is not advisable to operate
the matrix converter under over modulation for a long time but for a short
period, if demanded, for the ride-through operation.
provide unity input displacement factor, sinusoidal supply currents and load
voltages that were identical to the conventional matrix converters. These
topologies are referred as the “sparse matrix converters”. These sparse matrix
converters were classified as (i) Simple Sparse Matrix Converter (SSMC)
with 15 switches, (ii) Very Sparse Matrix Converter (VSMC) with 12
switches and (iii) Ultra Sparse Matrix Converter (USMC) with 9 switches
(Meng Yeong Lee 2009). The VSMC and the USMC were designed based on
the fact that the DC link current only flows in one direction. This constraint
makes the VSMC and the USMC not applicable for regenerative operation.
Meng Yeong Lee et al (2010), but was still able to generate multilevel output
waveforms. A detailed analysis of this topology was presented in Meng
Yeong Lee (2009).
The use of 3×6 matrix converters for a six phase induction machine
drive system was presented in Wang et al (2011), Ghalem and Azeddine
(2010). In addition to the changes in the topologies of the matrix converter,
the use of the polyphase matrix converter for innovative active generator was
demonstrated in Beguin (2012). It explained along with a prototype, the
commutation, the modulation and the control principles of a 27×3 matrix
converter.
based wind energy systems was carried out in Barakathi (2008). Control of
the reactive power supplied by a wind energy conversion system (WECS)
based on the induction generator fed by a matrix converter was presented in
Cardenas et al (2009). An increasing number of papers (Imayavaramban and
Wheeler 2007, Wheeler et al 2003 and Lillo 2006) investigating the
advantages/ limitations of the use of matrix converters in aircrafts are also
being reported.
The desire to use all the advantages offered by the matrix converter
has inspired me to work in this area for my PhD research. This research work
attempts to investigate the existing PWM techniques and suggest
modifications for improving the performance of the matrix converter. This
thesis focuses on devising easier methods to implement complex switching
strategies, study and mitigation of effects of the unbalance, topological
changes to increase the performance indices, proper use of the modulation
technique to eliminate the common mode voltage and a new direct torque
control procedure for the control of induction motor fed by the modified
matrix converter topology.
39
1.9 OBJECTIVES
cycles of the CMC from the available reference signals (without the need of
processors and memory) and simultaneously executes it with a simple digital
logic circuit making it suitable for online (computation-less) implementation.
Since the duty cycles of the fictitious converter inverter pair are combined
using a digital circuit to achieve the CMC duty cycle, the input current
harmonics increase. This increase in the input current harmonics is due to the
non-coordinated inverter and converter zero vectors. To improve the
performance of the DIDC technique, the carrier frequency adjustment method
is proposed to reduce the THD in the input current. The chapter includes
simulation and experimental validation to highlight the merits of the
approach.
simple dynamic ISVM approach for the matrix converter operation for the
unbalanced and the non-sinusoidal input voltage conditions are presented.
Analyses of the effects of the unbalance on the FDCB of the CMC operated
under ISVM PWM method is carried out. An unbalanced control method for
the CMC is developed which uses the line side switching functions
(converter) to track the oscillations of the average fictitious DC bus voltage
and generate a dynamic modulation index for the load side (inverter). This
approach is based on the simple compensation of the output voltage
modulation vector with respect to the oscillating fictitious DC bus vector. The
simulated results presented bring out the effectiveness of the proposed
technique.
Chapter 6 develops the DTC control method for the DTMC, which
uses the input phase voltage vectors (short vectors) and the input line voltage
vectors (long vectors). In the proposed algorithm, the large vectors are applied
during torque transition states whereas the short vectors are utilized for the
steady state conditions, which results in the reduction of the torque ripples.
43
CHAPTER 2
2.1 INTRODUCTION
The carrier based sine triangular PWM technique for the power
converters (choppers, inverters and multilevel inverters) is easy to implement
and widely used. The Space Vector PWM (SVPWM) techniques provide a
much higher voltage transfer ratio but require a processor and a memory for
implementation (Bose and Sutherland 1983). The comprehensive relation of
the two PWM methods provides a platform not only to transform from one to
the other, but also to develop different PWM modulators. Therefore, many
attempts have been made to unite and derive the relationship between these
two types of PWM methods in the voltage source inverters (Blasko 1997,
Holmes 1992, Bowes and Lai 1997).
simple but superior carrier based PWM technique for matrix converters, so
that these converters can be used in the industry. A simple PWM technique
called the Decoupled Indirect Duty Cycle (DIDC) PWM technique, developed
in this chapter, incorporates the carrier PWM technique with modifications.
The carrier based PWM technique developed in this chapter for the
CMC, reduces the complexity of the switching strategy and improves the
performance of the converter. The basic idea of the proposed DIDC PWM
technique is to reduce the computations required to calculate the duty cycle
over every switching period. The duty cycle information, extracted from the
input voltage signal and the output reference voltage signal, through a
classical analog circuit, directly generates the firing pulses. The proposed
carrier frequency adjustment method increases the performance of the DIDC
PWM technique and offers considerable improvement in the quality of the
output voltages and the input currents.
VDC+ VDC+
Sap Sbp Scp Sp
SAp SBp SCp Sp
VA Va
IA Ia
VB IB Ib Vb
VC IC Ic
Vc
SAn SBn SCn Sn San Sbn Scn Sn
VDC-
VDC -
µAa µAb µ Ac µI ap µI an
[ ][ ][
µ = µBa µBb µ Bc = µI bp µI bn ×
µCa µCb µCc µI cp µI cn
µCAp µCBp µCCp
µCAn µCBn µCCn ] (2.1)
Transformation matrix µ
Rf IA Lf RL LL Va Ia
VA
- +
VB IB Vb Ib
- +
IC Vc Ic
VC
- +
Equations (2.2) to (2.4) give the expected input currents at each leg
of the input converter.
I A = I m sin(ωs t + φi ) (2.2)
The duty cycle for a leg is defined as the time for which that leg
conducts from the source to load (independent of the direction of the current)
in a given sampling period. It is proportional to the absolute value of the input
reference, as given by Equations (2.5) to (2.7) and shown in Figure 2.3
DA = | sin ( ω s t + φ i ) | (2.5)
where, DA, DB and DC are the duty cycles of each leg, which are normalized
between 0 and 1.
X Y Z
Duty cycle
DB DA DC DB DA DC
Region 1-2 Region 3-4 Region 5-6 Region 7-8 Region 9-10 Regn 11-12
Time (s)
Table 2.1 Duty cycle variation in each leg of the input converter
where, μCAp, μCBp and μCCp are the modulation functions of the upper
(positive) switches corresponding to the phases A, B and C respectively;
μCAn, μCBn and μCCn are the modulation functions of the lower (negative)
switches corresponding to the phases A, B and C respectively.
VA IA VA IA VA IA
IB IB VB IB
VB VB
IC IC VC IC
VC VC
Figure 2.5 shows the modified duty cycle of the switches in the leg
A. When the phase A voltage is minimum, both the switches of the leg A
conduct for an additional period of (1 - D B) in the region 1 where the B phase
voltage is maximum, and conduct for an additional period of (1 - D C) in the
region 2 where the C phase voltage is maximum.
51
Duty cycle
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
DB DA DC
1-DB 1- DC
µCAp DA +(1-DB) DA DA DA DA DA + (1-DC)
Time (s)
Figure 2.5 Duty cycle formulations for the input converter switches S Ap
and SAn
Switches Regions
1 2 3 4 5 6
µCAp 1+DA-DB DA DA DA DA 1+ DA-DC
µCBp 0 0 0 1-DA 1+DB-DC DB
µCCp DC 1+DC-DB 1-DA 0 0 0
µCAn 1-DB 0 0 0 0 1-DC
µCBn DB DB DB 1+ DB-DA 1-DC 0
µCCn 0 1-DB 1+DC-DA DC DC DC
Switches Regions
7 8 9 10 11 12
µCAp 1-DB 0 0 0 0 1-DC
µCBp DB DB DB 1+ DB-DA 1-DC 0
µCCp 0 1-DB 1+DC-DA DC DC DC
µCAn 1+ DA-DB DA DA DA DA 1+DA-DC
µCBn 0 0 0 1-DA 1+DB-DC DB
µCCn DC 1+DC-DB 1-DA 0 0 0
52
0
Regions 1-2 Regions 3-4 Regions 5-6 Regions 7-8 Regions 9-10 Regn 11-12
Time (s)
Figure 2.6(a)Duty cycles of the positive switches for the input converter
1
Duty cycle
0
Regions 1-2 Regions 3-4 Regions 5-6 Regns 7-8 Regns 9-10 Regn 11-12
Time (s)
Figure 2.6(b) Duty cycles of the negative switches for the input converter
This section explains the carrier based PWM scheme for generating
the modulation function, for each switch, of the input converter. The input
phase voltages, the rectified input phase voltages and the triangular carrier
signal with frequency fsic, are processed using simple comparator circuits to
generate nine digital signals. The nine digital signals are (a) M 1 = A > B, (b)
M2 = B > C, (c) M3 = C > A, (d) PA = A > 0, (e) PB = B > 0, (f) PC = C > 0, (g)
LA = Trig < A, (h) LB = Trig < B and (i) LC = Trig < C, where PA, PB and PC
are the signals that indicate the polarity of the input phase voltages; L A, LB
53
and LC are the comparator signals that are high when the magnitude of the
triangular carrier wave is less than the magnitudes of the rectified input phase
voltages. The additionally generated digital signals M A = M1 . Ḿ 3, MB = M 2 . Ḿ 1,
and M C = M3 . Ḿ2 indicate the absolute maximum of the rectified input phase
voltages; N A = M3 . Ḿ1, N B = M 1 . Ḿ 2, and NC = M 2 . Ḿ 3 indicate the absolute
minimum of the rectified input phase voltages.
A digital circuit is designed to generate the six PWM pulses for the
input converter with the help of the signals Mi, Ni, Pi and Li where i=A, B, C
and the procedure to trigger the input converter is formulated as follows.
Let i be the phase with the maximum duty cycle, j be the phase with
the decreasing duty cycle and k be the phase with the increasing duty cycle.
If Mi=1, Li=1, and Pi=1, the switch Sip is ON; else if Mi=1,
Li=1 and Pi=0, then the switch Sin is ON.
If Lj=1 and Pj=1, the switch Sjp is ON; else if Lj=1 and Pj=0,
the switch Sjn is ON.
If Pk=1, Lj=0 and Li=1, the switch Skp is ON; else if Pk=0, Lj=0
and Li=1, the switch Skn is ON.
If Li=0 and Nj=1 then switches Sjp and Sjn are ON; else if Li=0
and Nk=1 then switches Skp and Skn are ON.
Signals Li, Mi, Ni, Pi and Table 2.1 are used to design the logic
circuit for triggering the input converter.
V a = V m sin ( ωo t + φ o ) (2.16)
The duty cycle for each leg is defined as the time for which the leg
conducts from the source to load (independent of the direction of current) in a
given sampling period. Equation (2.19) gives the leg duty cycles D a, Db and
Dc, implemented with the sine PWM method.
D a = D b = Dc =1
(2.19)
µI ip + µI in = 1 (2.20)
where, i = a, b, or c, and µIip and µIin are the duty cycles of the switches, in the
upper and the lower arms respectively, of the output converter.
Equations (2.21) to (2.23) give the duty cycle of each switch in the output
converter, implemented using the sine PWM method (Wang 2006).
µI ap = (1 + sin(ωo t + φ o) ) /2 (2.21)
V dc
it provides only as the peak value (Blasko 1996) of the locally averaged
2
phase voltage, which reduces the performance of the matrix converter. To
2
increase the DC bus utilization of the output converter by a factor , a third
√3
harmonic zero-sequence component (Houldsworth and Grant 1984 and
Holmes 1996) is used to modify the duty cycles of the legs. Equation (2.24)
gives the modified duty cycles of the legs
D a = D b = D c = 1 - Z3c (2.24)
where, Z3c is the third harmonic zero-sequence component whose shape and
offset is derived in Zhou and Wang (2002) and given by Equation (2.25)
where, K0 is the ratio that denotes the sharing between the two zero vectors
V0, V7 of the inverter. Except the points at which the duty cycles of the legs
are not 1, the output converter does not conduct current from the source to the
load for a time Z3c ×Ts. This means that all the legs of the output converter are
disconnected from the source for a period Z3c ×Ts. During the turn OFF of the
converter, a definite requirement is to provide a freewheeling path to the
inductive load currents. Figure 2.7 shows that this is achieved by turning ON
all the switches in the upper or the lower arms.
56
VDC+ VDC+
Sap Sbp Scp Sap Sbp Scp
Figure 2.7 Output converter freewheeling path for the load current
For reducing the common mode voltage in the matrix converter, the
zero vectors of the output converter are not significant, as the common mode
voltage depends only on the zero vectors of the input converter. Hence, for
ease of digital implementation, the value K 0 is chosen as 1, i.e., the vector V 7
is used in the output converter. This modifies Equation (2.25) as
Equation (2.26)
where, µImax_p = max |µIa, µIb, µIc|. Figure 2.8 shows the modified duty cycles
of the legs of the output converter.
1
Duty cycle
Da = Db = Dc = 1- Z3c
Time (s)
57
2
µI *ap = ((1 + sin(ω o t + φ o) ) /2) + Z3c (2.27)
√3
2
µI *bp = ((1 + sin (ω o t + φo - 1200 ) )/2) + Z3c (2.28)
√3
2
µI *cp = ((1 + sin( ω o t + φo + 1200 ))/2) + Z3c (2.29)
√3
µ ap µ ap µ cp
1
µ*ap
Duty cycle
Z3c
Time (s)
Figure 2.9 Formulation of switch duty cycle, S ap, for the output
converter
0
Time (s)
Figure 2.10(a) Duty cycles of the positive switches for the output
converter
1
Duty cycle
0
Time (s)
Figure 2.10(b) Duty cycles of the negative switches for the output
converter
W3+ & W1+ & W2- W2+ & W3- W3+ &
W1- W1-
Time (s)
59
This defines the duty cycle for each switch, as a function of the new
reference signals |W123|, in each distinct region, as given in Table 2.3. The new
reference signals are used in the digital implementation for generating the
switching signals of the output converter.
Table 2.3 Duty cycles for all the switches of the output converter as a
function of |W123|
Regions W1+ & W2- W2+ & W3- W3+ & W1-
Switches
µIap 1 (1- |W2|) (1- |W1|)
µIbp (1- |W2|) 1 (1- |W3|)
µIcp (1- |W1|) (1- |W3|) 1
µIan 0 |W2| |W1|
µIbn |W2| 0 |W3|
µIcn |W1| |W3| 0
2.5.1 Carrier Based Implementation Procedure for the Output
Converter
voltage is less than the rectified output phase voltages. The pulse widths of
the signals L1, L2 and L3 are proportional to |W1|, |W2|, and |W3| respectively;
the pulse widths of the signals Ĺ1 , L´ 2 , and Ĺ3 are proportional to |1 - W1|, |1 -
W2|, and |1 - W3| respectively.
Finally, the switching signals for the matrix converter are extracted
from the logic circuit that combines the information of the input converter and
the output converter. The circuit generates the switching signals for the nine
switches of the conventional matrix converter, as given by Equation (2.30)
(Cha 2004).
SAa SAb SAc Sap . SAp + San . SAn Sap .SBp + San .S Bn Sap .SCp + San .SCn
[ ][
SBa SBb SBc = S bp . SAp + Sbn .S An Sbp .SBp + Sbn .SBn S bp .SCp + Sbn .SCn
SCa SCb SCc Scp . SAp + Scn . SAn Scp .SBp + Scn .SBn Scp .SCp + Scn .SCn ]
(2.30)
50 µs. The proposed method requires a set of logic gates and multiplexer
circuits to generate the duty cycles, which reduces the time for generating the
firing pulses to less than 5 µs. Thus, the proposed method permits the control
of a matrix converter at higher switching frequencies than those of the
Venturini method.
100
THD =
1
I1 √∑
i=2
I2i (2.31)
where, I1 and Ii are respectively the fundamental and ith harmonic components
of the output currents. The THDs for different values of fs ic and fsoc, when the
load current frequency is 20 Hz, is shown in Figure 2.12(b).
62
5.7
5.5
5.4
5.3
5.2
4 6 8 10 12 14 16 18 20
Frequency (kHz)
6
THD % of Fundamental
0
20 18 16 14 18 20
12 14 16
10 8 10 12
6 6 8
4 4
FCSC Carrier Frequency(kHz) FVSI Carrier Frequency(kHz)
The THD for various ratios of fsic and fsoc for varying load
frequencies is plotted in Figure 2.13(a). It is observed that the THD for all the
output frequencies is nearly the same value of 3.14%, for the ratio of input
and output carrier frequencies of 1:3/4, which is well within the acceptable
limits. Hence, to improve the THD of the input currents for a wide range of
output frequencies, in the proposed method, the switching frequencies of the
input and the output converters are chosen to be in the ratio of 1: 3/4.
The THD vs. output load current frequencies for equal input and
output carrier frequencies and THD vs. output load current frequencies for the
input and the output carrier frequency in the ratio of 1: 3/4, are given in
Figure 2.13(b).
63
Figure 2.13(a) THD for different ratios of fsic and fsoc with varying fo
Figure 2.13(b) THD for ratios of fsic= fsoc and 3/4 fsic= fsoc with varying fo
2.8 SIMULATION
Quantity Value
R-L Load R = 20 Ω , L = 21 mH
Input Phase Voltage 100 V
Input Voltage Frequency 50 Hz
Input Filter L = 2.5 mH, C = 10 µF, Rd = 15 Ω
Output Voltage Frequency 25 Hz
Modulation Index 0.75
Switching frequency 7 kHz
5
Mag (% of Fundamental)
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency (Hz)
(a)
65
5
Fundamental (50Hz) = 32.75 , THD= 1.84%
4
Mag (% of Fundamental)
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency (Hz)
(b)
5
Mag (% of Fundamental)
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency (Hz)
Figures 2.16(a) and 2.16(b) illustrate the input and the output
current waveforms for unity input power factor. It is seen that the filtered input
and output currents are nearly sinusoidal. Thus, the proposed method provides
high quality input and output currents. For a modulation index of 0.75, the
output current spectrum has a THD of 0.56% and the filtered input current
spectrum has a THD of 1.84%, as shown in Figures 2.15 and 2.14 (b). For a
lower modulation index, the method shows increased harmonic contents in the
input current spectrum while the harmonic content in the output current
66
spectrum remains nearly the same. Figures 2.16(c) and 2.16(d) show the
output line and phase voltage.
2
IABC(A)
-2
-4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(a)
2
Iabc(A)
-2
-4
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(b)
200
100
Vab(V)
-100
-200
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(c)
67
200
100
Van(V)
-100
-200
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(d)
Figure 2.16 Simulation results (a) input current, (b) output current,
(c) output line voltage and (d) output phase voltage
50 Hz Bipolar Sine
Wave Reference Zero Crossing Load Current
Signals Detectors direction sensor
3 3
3
3 6SD106EI
Unipolar High Precision
Frequency Carrier CONCEPT
Rectifier FPGA
(kHz) 3 Driver Module
Module 9
Comparators
3 9
3 3
Matrix Converter
Module
(a)
70
(b)
(c)
(d)
Figure 2.19 Hardware results (a) input current, (b) output current,
(c) output line voltage and (d) output phase voltage
2.10 SUMMARY
A new method was proposed to extract the duty cycle for each
switch, under unity input power factor, using the carrier based PWM
technique for the CMC. This technique avoids the need for computing the
duty cycles for switches over every sampling period. Furthermore, the
71
method eliminates the need for ADCs, and high-speed processors, which are
required in the ISVM. In addition, the proposed method does not affect the
voltage transfer ratio.
CHAPTER 3
3.1 INTRODUCTION
A space vector based scheme to reduce the CMV in the cascaded multilevel
inverters, has been proposed in Gupta and Khambadkone (2007). Kanchan et
al (2006) have proposed the elimination of the CMV for open-ended winding
based induction motor drives fed by a three-level inverter. Gupta et al (2010)
have proposed the same for a matrix converter. However, methods to
eliminate the common mode voltage in drives fed by a CMC without an open-
ended winding are not available in the literature.
Va
Cws
SOURCE
Cwr
Crs
Vb
Cwr Cwr
Vc
Cws
dI a
V a - V cm = R Ia + L (3.1)
dt
dI b
V b - V cm = R I b + L (3.2)
dt
dI c
V c - V cm = R I c + L (3.3)
dt
Vcm = ( V a + Vb + Vc ) /3 (3.4)
The CMC topology, shown in Figure 1.2, has been chosen in this
work, as it is the only possible converter topology where the CMV can be
eliminated. Equation (3.5) gives the expression for the source neutral voltage
VN that is zero for any three-phase three-wire balanced system.
VN = ( V A + V B + V C ) /3 (3.5)
From Equations (3.4) and (3.5), the sum of load voltages Va, Vb and
Vc should be equal to the sum of input phase voltages VA, VB and VC for the
load neutral voltage Vn to be equal to VN i.e. zero. Hence, the elimination of
the CMV introduces a third constraint in addition to the two constraints
explained in Equation (1.2), namely that at any given instant, all the three
input phases must be connected to the load. This additional constraint, used
for the elimination of the CMV, allows the use of only six switching vectors
that connects all the input phases to the output phases, as shown in
Figures 3.2(a) and 3.2(b), from the valid 27 vectors of the CMC. The
76
remaining vectors that are not used are the 18 stationary vectors and the three
zero vectors, as shown in Figure 3.2(c). These six vectors used are termed as
the rotating space vectors since their positions in the space are not fixed. They
are also called as ZCMVV since they produce zero CMV. Among these six
rotating space vectors, three space vectors rotate in the direction of the output
frame and the remaining three space vectors rotate in the direction opposite to
the output reference frame.
Y Y
V7 ωs V4 ωs
V9 V6
X X
V8 V5
(a) (b)
V14,V17,V25 Y V10,V19,V26
V15,V16,V24 V13,V21,V22
V1, V2 ,V3 X
(c)
V11,V18,V27 V12,V20,V23
Figure 3.3(a) shows the output space vectors, when any of the three
possible switching patterns, given in Equation (3.6), are employed.
3
v abc+ = V i e jω t s
(3.7)
2
2π
3 j(ω t + )
v cab+ =
s
3
Vi e (3.8)
2
2π
3 j(ω t- )
v bca + =
s
3
Vi e (3.9)
2
3
vo = V o e jω t o
(3.10)
2
where, v abc + , v cab+ , v bca + are the active vectors with magnitude V i and vo is the
reference output vector with magnitude Vo.
Y
Y
vcab+ vcab+
(ωo - ωs)
VREF
ωo vabc+ VREF )
θv = (θ0 - θs) X
θS ωs X
θO vabc+
Mi_max=0.5
Figure 3.3 Dynamic space vector PWM (a) positive sequence rotating
reference frame and (b) fixed reference
Equations (3.11) to (3.14) give the duty cycles of the active vectors
and the zero vectors obtained by the sine law of triangles, as shown in
Figure 3.4.
d α vα dβ v β V o(REF)
o
= = (3.11)
sin ( 120 - θ v ) sin ( θv ) sin ( 60o )
d β = m v sin ( θv ) (3.13)
d0= 1 - dα - dβ (3.14)
y
ωo vβ
vβ
Vo(REF)
120˚ Vo(REF)
θv 120˚-θv
ωs x
60˚ dβvβ
θv
vα
(a)
dαvα vα
(b)
Figure 3.4 Duty cycle calculation (a) sector 1 and (b) weighted
combination of the active vectors
the switching time of the zero vector, the RSVM technique uses all the three
active vectors with equal duty ratios. Equations (3.15) to (3.17) give the
modified duty ratios for the RSVM technique.
+ ' o d0
d 1 =d α = m v sin (120 - θv )+ (3.15)
3
+ ' d0
d 2 = dβ = m v sin ( θv ) + (3.16)
3
+ ' d0
d 3 = d0 = (3.17)
3
v o = d 1+ v1 + d2+ v 2 + d 3+ v 3 (3.18)
d 1+ +d 2+ + d3+ = 1 (3.19)
where, d1+, d2+, d3+ are the duty cycles and v1, v2, v3 are the active positive
rotating vectors respectively.
Table 3.1 Positive rotating switching vectors and sector no. for the CMC
1 2 mv
D Aa = DBb = D Cc = + cos (ω o t - ωs t) (3.20)
3 3
1 2 mv 2π
D Ba = D Cb = D Ac = + cos ((ωo t - ωs t) + )
3 3 3
(3.21)
1 2 mv 2π
D Ca = D Ab = D Bc = + cos ((ωo t - ωs t) - )
3 3 3
(3.22)
Table 3.2 Negative rotating switching vectors and sector no. for the CMC
+ 3 j (ω t-ρ)
i abc = I e o
(3.33)
2 o
2π
+ 3 j (ω t-ρ +o
3
)
i cab = Io e (3.34)
2
2π
+ 3 j(ω t-ρ-
3o )
i bca = Io e (3.35)
2
3
is = I e j(ω t-ρ)
s
(3.36)
2 o
3
i acb- = Io e - j (ω t-ρ)
o
(3.37)
2
2π
3 -j (ω t-ρ + )
i bac- =
o
3
Io e (3.38)
2
2π
3 - j(ω t-ρ- )
i cba - =
o
3
Io e (3.39)
2
3
is = I o e j(ω t+ρ)
s
(3.40)
2
82
Equations (3.36) and (3.40) show that the input current lags or leads
respectively the input voltage by ρ degrees when positive or negative
directional voltage vectors are applied, where cos ρ lagging is the output load
power factor, as shown in Figure 3.5(a). Since the output power factor
depends upon the load, it is not possible to control the output power factor.
Hence, the only way to control the input power factor cos is to apply both
+ve and -ve rotating voltage space vectors at the output terminals, as shown in
Figures 3.5(b) to 3.5(d). Equation (3.50) gives the ratio r with which the +ve
and –ve rotating voltage space vectors are applied to decide the input power
factor. However, the input power factor can be controlled in a limited range
between the output power factor and unity i.e. cos ρ < cos < 1.
Equations (3.41) to (3.49) give the duty cycles of each switch.
1 2r m v 2(1-r) m v
D Aa = + cos ( ωo t- ωs t ) + cos ( ω o t+ωs t ) (3.41)
3 3 3
1 2r m v 2π 2(1-r) m v 2π
D Ba = + cos ( ( ω o t- ωs t ) + )+ cos ( ( ω o t+ ωs t ) - ) (3.42)
3 3 3 3 3
1 2r m v 2π 2(1-r) m v 2π
D Ca = + cos ( ( ωo t-ω s t ) - )+ cos ( ( ωo t+ω s t ) + ) (3.43)
3 3 3 3 3
1 2r m v 2(1-r) m v 2π
D Bb = + cos ( ω o t- ωs t ) + cos ( ( ωo t+ω s t ) + ) (3.44)
3 3 3 3
1 2r m v 2π 2(1-r) m v
D Cb = + cos ( ( ω o t- ω s t ) + )+ cos ( ωo t+ω s t ) (3.45)
3 3 3 3
1 2r m v 2π 2(1-r) m v 2π
D Ab = + cos ( ( ω o t- ωs t ) - )+ cos ( ( ω o t+ ωs t ) - ) (3.46)
3 3 3 3 3
1 2r m v 2(1-r) m v 2π
D Cc = + cos ( ωo t- ωs t ) + cos ( ( ωo t+ω s t ) - ) (3.47)
3 3 3 3
83
1 2r m v 2π 2(1-r) m v 2π
D Ac = + cos ( ( ω o t- ω s t ) + )+ cos ( ( ωo t+ ωs t ) + ) (3.48)
3 3 3 3 3
1 2r m v 2π 2(1-r) m v
D Bc = + cos ( ( ωo t-ω s t ) - )+ cos ( ω o t+ ωs t ) (3.49)
3 3 3 3
y
Vs ωs
Is+
Is- ωs ρ
ρ ω0
Vo (REF)
ρ ω0
Io x (a)
ωs ωs ωs
Is+ Is+ Is+
d+ Is+
d+ Is+ d+ Is+
ρ Is θ
ρ θ ρ ρ
Vs Vs Vs
d- Is- Is d- Is- d- Is- Is
Figure 3.5 Input power factor control (a) input current positions,
(b) lagging power factor (d+ < d-), (c) unity power factor (d+
= d-) and (d) leading power factor (d- < d+)
Let T+ be the time for which the positive rotating space vectors are
applied and T- be the time for which the negative rotating space vectors are
applied, within a given sampling time Ts. The respective duty ratios are d+ and
84
d- and they satisfy the relation d + + d- =1 at all times. Equation (3.50) gives
the input power angle θ.
The RSVM technique for the CMC uses only three vectors with the
modulation index limited to 0.5, which is a major limitation of the RSVM
technique. The shortcoming of this technique is overcome by using a 6×3
CMC consisting of two 3×3 CMCs [MC x , MC y] fed by a three-phase center-
tapped transformer. The transformer produces 180° shifted space vector
pattern by generating a six-phase supply, as shown in Figure 3.7. In the
proposed topology, the modulation index is extended to 0.866 with the help of
the newly available three space vectors, as shown in Figure 3.6.
Equations (3.51) and (3.52) give the six switching patterns for the counter-
clockwise rotating space vectors, where all the elements of the SMC_x are equal
to 0 if any one switching sequence of the SMC_y is applied. Similarly, all the
elements of the SMC_y are equal to 0 if any one switching sequence of the SMC_x
is applied.
85
Y
vcab+x vbca+y
ωo
VREF
θv X
vabc+y ωs vabc+x
Mi_max=0.866
vbca+x vcab+y
(a)
Y
vbac-x vcba-y
ωo
VREF
θv X
vacb-y ωs ωs vacb-x
Mi_max=0.866
vcba-x vbac-y
(b)
Figure 3.6 Space vector distribution of the PSDSMC (a) +ve sequence
vectors and (b) -ve sequence vectors
86
Ia Ib Ic
[
= SBax SBbx SBcx
SCax SCbx SCcx ]
1 0 0 0 1 0 0 0 1
[ ] [ ] [ ]
= 0 1 0 or 0 0 1 or 1 0 0
0 0 1 1 0 0 0 1 0
(3.51)
[
= SBay SBby SBcy
SCay SCby SCcy ]
1 0 0 0 1 0 0 0 1
[ ] [ ] [ ]
= 0 1 0 or 0 0 1 or 1 0 0
0 0 1 1 0 0 0 1 0
(3.52)
87
3
v abc+ x = V i e jω ts
(3.53)
2
2π
+ 3 j(ω t +
3 s )
v cab x = Vi e (3.54)
2
2π
+ 3 j(ω t-
3 s )
v bca x = Vi e (3.55)
2
3
v abc+ y = V e j(ω t+π) s
(3.56)
2 i
π
3 j(ω t- )
v cab+ y =
s
3
Vi e (3.57)
2
π
+ 3 j(ω t +
3 s )
v bca y = Vi e (3.58)
2
3
vo = V o e jω t
0
(3.59)
2
where, v abc+ x , v cab+ x , v bca + x and v abc+ y , v cab+ y , v bca + y are respectively the
active positive rotating switching vectors corresponding to SMC_x and SMC_y
respectively. Vi and Vo are the magnitudes of the input and the output space
vectors respectively. The duty cycles of the active vectors and the zero vector
are computed using the sine law of triangles, as explained in section 2, and
given by Equation (3.60).
technique utilizes two opposite active vectors in equal ratio within the time
for zero switching that modifies the duty ratios, as given by Equation (3.61).
+ + d0 + d0
d 1 =d α , d2 = dβ + , d3 = (3.61)
2 2
v o = d1+ v 1 + d 2+ v 2 + d 3+ v 3 (3.62)
Table 3.3 Positive rotating switching vectors and sector no. for the
PSDSMC
Table 3.4 Negative rotating switching vectors and sector no. for the
PSDSMC
It is also observed that the input power factor control reduces the
magnitude of the input current, as shown in Figure 3.8, by a factor given in
Equation (3.63)
|I s_c| cos ρ
+
= (3.63)
|I | s cos ( tan -1 [ ( 1-2 d+ ) tan ρ ])
where, |Is_c| and | Is+| are respectively the peak magnitude of the controlled
input current and the peak magnitude of the uncontrolled input current of the
CMC.
ωs
V Is+
Magnitude
d+ Is+
I+s ρ Is_c
I-s
Is_c ρ Vs
d- Is-
- ρ-- ρ- Time (s)
Is- (b)
(a)
cos ρ (3.64)
m v_c = mv
-1 +
cos ( tan [ ( 1-2 d ) tan ρ ] )
From Equation (3.5), it can be seen that when unbalanced inputs are
applied to the CMC or the PSDSMC, the RSVM technique does not eliminate
the CMV. For the inputs containing homopolar harmonics, the RSVM
technique fails to eliminate the CMV, as homopolar harmonics introduce a
zero sequence component. However, the magnitudes of the CMV introduced
under such conditions are very low and that they do not affect the system very
seriously. The RSVM technique is unaffected by non-homopolar harmonics
since they do not introduce a zero sequence component in the system.
3.9 SIMULATION
Simulation of the CMC and the proposed PSDSMC was carried out
using mathematical models, as shown in Figures 3.9 and 3.10, and also
verified using ideal switches. The system parameters used in the simulation
were: Supply – 220 V, 50 Hz, Load – R=5Ω, L=12 mH, cos ρ = 0.8 at 25 Hz
output frequency and switching frequency of 7 kHz. The input filter
capacitance and inductance were designed to be C f =10 μF and Lf =1 mH with
a damping resistor rd = 15Ω for filtering the higher order frequencies very
near to the switching frequency.
91
Ia
VA IA
VaN
IC
Figure 3.9 Mathematical model of the CMC (a) voltage model and
(b) current model
SACVR_1
VA
VaN
VA’
VB’
VcN
VC’
IA
Ia
IA’
IB
Ib
IB’
Ic
IC
IC’
Figure 3.10 Mathematical model of the PSDSMC (a) voltage model and
(b) current model
Figures 3.12 and 3.13 show the simulation results of the CMC
controlled by the RSVM technique without and with current control
respectively. Figures 3.14 and 3.15 show the simulation results of the
PSDSMC controlled by the RSVM technique without and with current control
respectively. Figures 3.3 and 3.6 show that the modulation index of the
PSDSMC controlled by the RSVM technique increases by 73.2% as
compared to modulation index of the CMC controlled by the RSVM
technique.
93
Vector Selection
Timing V1+, V2+, V3+
Vectors
Matching Vector Selection
V1-, V2-, V3-
Switching Pulses
40
V A /8 (V ) , IA (A )
20
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(a)
Figure 3.12 (Continued)
94
400
Vab(V) 200
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(b)
400
200
Va(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(c)
400
200
Vab(V)
-200
-400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time(s) x 10
-3
(d)
20
IA(A), IB(A), IC(A)
10
-10
-20
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(e)
Figure 3.12 (Continued)
95
20
-10
-20
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(f)
Figure 3.12 RSVM technique for the CMC without current control (a)
input current and voltage respectively, (b) output line
voltage, (c) output phase voltage, (d) output line voltage
magnified, (e) input currents and (f) output currents
40
V A /8(V ) , IA (A )
20
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(a)
400
200
Vab(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(b)
400
200
Va(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(c)
Figure 3.13 (Continued)
96
400
200
Vab(V)
0
-200
-400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time(s) x 10
-3
(d)
20
IA(A), IB(A), IC(A)
10
-10
-20
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(e)
20
Ia(A), Ib(A), Ic(A)
10
-10
-20
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(f)
Figure 3.13 RSVM technique for the CMC with current control (a)
input current and voltage respectively, (b) output line
voltage, (c) output phase voltage, (d) output line voltage
magnified, (e) input currents and (f) output currents
Figures 3.12(d), 3.13(d), 3.14(d) and 3.15(d) show that the output
voltage consists of three levels when the positive rotating vectors are applied
while it consists of six levels when both the positive and the negative vectors
are applied. This indicates that the number of switching states increases with
97
the current control technique. Figures 3.12(a), 3.13(a), 3.14(a) and 3.15(a)
show that the input peak current reduces approximately by the factor given by
the Equation (3.30).
100
VA/4 (V), IA(A)
50
-50
-100
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(a)
400
200
Vab(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(b)
400
200
Va(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(c)
400
200
V ab (V )
0
-200
-400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time(s) x 10
-3
(d)
40
20
IA(A), IB(A), IC(A)
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(e)
40
Ia(A), Ib(A), Ic(A)
20
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(f)
Figure 3.14 RSVM technique for the PSDSMC without current control
(a) input current and voltage respectively, (b) output line
voltage, (c) output phase voltage, (d) output line voltage
Magnified, (e) input currents and (f) output currents
99
100
50
VA/4 (V) , IA(A)
-50
-100
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(a)
400
200
Vab(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(b)
400
200
Va(V)
-200
-400
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(c)
400
200
Vab(V)
0
-200
-400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time(s) x 10
-3
(d)
40
IA(A), IB(A), IC(A)
20
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(e)
40
Ia(A), Ib(A), Ic(A)
20
-20
-40
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Time(s)
(f)
Figure 3.15 RSVM technique for the PSDSMC with current control
(a) input current and voltage respectively, (b) output line
voltage, (c) output phase voltage, (d) output line voltage
magnified, (e) input currents and (f) output currents
From Figures 3.3 and 3.6, it can be observed that for the CMC and
the PSDSMC the space vectors are distributed by 120 o and 60o respectively.
Due to this, the maximum voltage stresses on the devices during the
101
commutation for the CMC and the PSDSMC are √3*Vm and Vm respectively.
Hence, it can be inferred that during commutation the devices in the CMC
topology are subjected to higher voltage stresses.
Figure 3.16 (a) shows the CMV induced due to the ISVM technique
in the CMC and Figure 3.16 (b) shows the elimination of the CMV by the
RSVM technique for both the CMC and the PSDSMC. The peak of the CMV
can be as high as the magnitude of the input phase voltage in the ISVM
technique. This has been eliminated in the proposed RSVM technique along
with the input current control.
200
Vcm (V)
-200
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
Time(s)
(a)
300
200
100
Vcm(V)
0
-100
-200
-300
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
Time(s)
(b)
Figure 3.16 Common mode voltage (a) ISVM technique for the CMC
and (b) RSVM technique for the CMC (or) the PSDSMC
102
200
100
VABC (V)
0
-100
-200
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Time (s)
(a)
100
50
Vcm (V)
-50
-100
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Time (s)
(b)
100
50
Vcm (V)
-50
-100
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
Time (s)
(c)
Figure 3.17 (a) Unbalanced input voltage, (b) CMV of the CMC and
(c) CMV of the PSDSMC
103
200
V A B C (V )
100
-100
-200
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Time (s)
(a)
100
50
Vcm (V)
-50
-100
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Time (s)
(b)
100
50
Vcm (V)
-50
-100
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Time (s)
(c)
Figure 3.18 (a) Non-sinusoidal input voltage, (b) CMV of the CMC and
(c) CMV of the PSDSMC
104
3.10 SUMMARY
The input current control range of the PSDSMC and the CMC is
limited from unity power factor to output power factor. In the current
controlled technique for the PSDSMC and the CMC, the modulation index
reduces further. It is also found that the switching stresses of the individual
switches reduce in the PSDSMC as compared to the CMC because of the
additional states introduced in the PSDSMC topology. Hence, the proposed
PSDSMC modulated by the RSVM technique can be used for the elimination
of the CMV in electrical machines with a higher modulation index.
105
CHAPTER 4
4.1 INTRODUCTION
under unbalanced input voltage have been reported in the literature (Casadei
et al 1998, Blaabjerg et al 2002, Zhang et al 2001 and Sunter et al 2002).
However, a simple fictitious DC bus (FDCB) based analysis of the unbalance
and the technique for mitigating the effects of the unbalance has not been
reported for the CMC. An unbalance control method for the CMC is
developed in this chapter. The output of the converter remains balanced with
the input current being distorted.
T
sin ( ω o t+ φo ) sin ( ωs t- φi )
D= m c m v
[ ][ ]
( 2π
3 )
sin ωo t+φ o -
2π
sin ( ω t+φ + )
o
3 o
× ( 2π3 )
sin ωs t- φ i -
s
2π
sin ( ω t- φ + )
3 i
(4.1)
T
sin ( ωs t- φi ) Vam sin ( ω s t )
VDC = m c
[ ][ ]
( 2π3 )
sin ω s t- φ i -
2π
sin ( ω t- φ + )
s i
3
×
Vcm
( 2π3 )
V bm sin ωs t-
s
2π
sin ( ω t+ )
3
(4.2)
that are unbalanced, the Fourier series for the phase voltages can be expressed
by Equation (4.4).
m
Va = ∑ Va,i sin( iωs t+ φi ) (4.4)
i=1
m
V pi sin ( iωs t+φ i + ϕ pi ) + V ni sin ( iωs t+φ i +ϕ ni )
Va = ∑
i=1 [ + V 0i sin ( iωs t+ φi + ϕ0i ) ] (4.5)
The current lags the voltage by ρ for all the phases. For the ith
harmonic with positive sequence Vpi ∠ ( φi + ϕpi ) and negative sequence
V ni ∠ ( φi + ϕni ), the space vector of the current for positive sequence of the ith
3
f i+ = f ai+ + f bi+ + f ci+ = V pi ( cos ( ( i-1) ω s t+ ϕpi +φ i +ρ ) ) (4.9)
2
Similarly, the space vector of the current for the negative sequence
of the ith harmonic is given by Equation (4.10).
3
f -i = f -ai + f -bi + f -ci = - V cos ( ( i+1 ) ωs t+ ϕni + φi -ρ ) ) (4.10)
2 ni (
f i = f i+ + f -i (4.11)
109
3
V DC_Calc = ( V cos ( ϕ p1 +ρ ) + Vn1 sin ( 2 ωs t + ϕ n1 -ρ- 900 ))
2 p1
(4.13)
Time (s)
(a) (b)
VO
LT
AG
E
(V)
TIME (s)
3
I DC_Calc = ( I p1 cos ( ϕ p1 +ρ ) + In1 sin ( 2 ωs t + ϕn1 - ρ - 900 ) ) (4.17)
2
I in = I f + I h (4.18)
Figure 4.3 shows the input line voltage space vector and the
rectifier current space vector. Equation (4.21) gives the synthesized FDCB
voltage.
VLM ejπ/2
IM ejπ/2
I3 [01-1]
IM ejπ/6
IM ej5π/6 2 1 I2 [10-1]
I4 [-110]
3 S3 0
S2
IM e-j5π/6 IM e-jπ/6
S1 I1 [1-10]
I5 [-101]
4 5
VLM ej5π/6 VLM ejπ/6
IM e-jπ/2
I6 [0-11]
Figure 4.3 The input voltage space vector and the rectifier current
space vector
1 S1 0, 3
2 S2 1, 4
3 S3 5, 2
where, V1, V2 are the line voltages of the corresponding sector and d Iα, d Iβ are
the duty cycle of the input current space vector with m c =1. The output
modulation index is modified dynamically based on the computed FDCB
voltage. Vdc_min is chosen as the maximum length of output voltage vectors.
For an increase in the FDCB voltage above this limit, the modulation index of
the voltage source converter is reduced correspondingly to keep the output
vector at the constant value as given in Equation (4.22). The output voltage
113
vector traces a circular path; hence, a balanced output voltage is obtained for
the three-phases. A simple memory technique is used to find the Vdc_min in
every cycle and this value is used in the subsequent cycle. As a result, the
harmonic characteristic of the output current is improved while the input
current harmonics are left uncompensated
VA = V m sin ( ωs t )
2π
(
V B =(1 - b)× V m sin ωs t-
3 )
2π
(
V C =(1 - a)× V m sin ωs t+
3 ) (4.23)
where, 0< a, b <1 are the unbalance constants in each phase. Then from Kang
and Sul (1997), the positive sequence voltage of each phase can be obtained
as given in Equation (4.24).
1 2T
VpA =
3 (
VA ( t ) + VB t - (
3 ) ( ))=(1- a3 - b3 ) × V
+ VC t -
T
3 A
114
1 2T
V pB =
3 (
VB ( t ) + V C t -(3 ) ( )) =(1- a3 - 3b ) × V
+ VA t -
T
3 B
1 2T
V pC =
3 (
VC(t )+ VA t -
3 (
+ VB t -
T
3) ( )) =(1- a3 - 3b ) × V C (4.24)
a b ⃗
⃗
(
VP = 1 - -
3 3
× V in ) ( 00 ) (4.25)
1
VnA =
3 ( T
( ) ( )) = - 13 ( aV + bV )
VA ( t ) + VB t- + VC t-
3
2T
3 B C
1
VnB =
( T
( ) ( )) = - 13 (aV + bV )
V ( t ) + VC t- + V A t-
3 B 3
2T
3 A B
1
VnC =
3 ( T
3( ) ( )) = - 13 (aV + bV )
VC ( t ) + V A t- + V B t-
2T
3 C A (4.26)
1
V in 1200 + b× ⃗
V N = - ( a× ⃗
⃗ V in -1200 )
3
V N| = -
|⃗ √a 2 + b2 -ab
3
(4.27)
V N| √ a 2 + b 2 -ab
|⃗
= (4.28)
V P|
|⃗ ( a+b ) -3
v 2n v 2p
K ub
max
( )|
=0.866× 1-
v 2p
×
v in|
2 (4.29)
2
a 2 + b2 -ab (3-(a+b) )
K ub
max
(
=0.866× 1-
( a+b-3 )2
× 2
3 ) (4.30)
Figure 4.4 Voltage transfer ratio vs unbalance factor across two phases
116
From Equation (4.3), it can be seen that as the input power factor
decreases, the fictitious DC bus voltage decreases. Hence, for a balanced
input, the output voltage transfer ratio decreases as the power factor
decreases. Using the same idea, it can be written that the maximum voltage
transfer ratio Kp ub
max for an unbalanced system with input power factor control
2
a 2 + b 2 -ab (3-(a+b) )
Kp ub
(
max =0.866× 1-
( a+b-3 )2) ×
32
× cos ( φ i ) ( 4.31 )
the input power factor and the unbalance ratio a = 0, 0>b >1. It is hence
preferred to operate the CMC at unity power factor at the input during higher
input unbalance to improve the voltage transfer ratio. This avoids the derating
of the load to the largest possible extent.
Figure 4.5 Voltage transfer ratio vs unbalance factor and input power
factor
4.8 SIMULATION
117
0
0 50 100 150 200 250 300 350 400 450 500 550
Frequency (Hz)
(a)
Fundamental (50Hz) = 22.88 , THD= 6.25%
10
Mag (% of Fundamental)
0
0 50 100 150 200 250 300 350 400 450 500 550
Frequency (Hz)
(b)
118
0
0 100 200 300 400 500 600
Frequency (Hz)
(c)
Fundamental (25Hz) = 32.39 , THD= 0.70%
10
Mag (% of Fundamental)
0
0 100 200 300 400 500 600
Frequency (Hz)
(d)
Figure 4.6 Harmonic spectrum (a) uncompensated input current,
(b) compensated input current, (c) uncompensated output
current and (d) compensated output current
Table 4.2 Simulation parameters for the unbalance control of the CMC
Quantity Value
R-L Load RL = 2.7Ω , LL = 4.77 mH
Input phase voltage 100 V
Input voltage frequency 50 Hz
Input filter L = 1 mH, C = 35 µF, Rd =15Ω
Output Voltage frequency 40 Hz
Switching frequency 6 kHz
Unbalance & Harmonic V a =100 % V b =90 % V c =100 %
Content V a 2=4 % V b 2=10 % V c 2=10 %
V a 3=3 % V b 3=25 % V c 3=12 %
The input voltages were initially kept balanced. The unbalance and
harmonics, given in Table 4.2, were created at the source at 0.1 ms.
Oscillations in the FDCB, as shown in Figure 4.8(a), reflect harmonics and
unbalance in the input. By modifying the modulation index dynamically, as
shown in the Figure 4.7, it is found that the output current spectrum contains
no harmonic content and is balanced with THD = 0.73%. The input current
spectrum shows harmonic contents with THD =11.19% with a large third
harmonic component of 8.7%, as shown in Figure 4.8.
(a) (b)
Figure 4.7 Polar plot of the modified mv,comp for (a) f=20 Hz, (b) f=40 Hz
120
150
100
50
0
0.05 0.1 0.15 0.2
Time(s)
(a)
Source Voltage sV[ABC] (V)
100
50
-50
-100
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time(s)
(b)
150
100
Load Voltage (V)
50
0
-50
-100
-150
0.05 0.1 0.15 0.2
Time(s)
(c)
30
t I B C ] (A )
20
10
So u rce C u rreSn [A
-10
-20
-30
0.05 0.1 0.15 0.2
Time
Time(s) (s)
(d)
40
-20
-40
0.05 0.1 0.15 0.2
Time(s)
(e)
0
0 50 100 150 200 250 300 350 400 450 500 550
Frequency (Hz)
(f)
0
0 40 80 120 160 200 240 280 320 360 400 440 480 520 560
Frequency (Hz)
(g)
Figure 4.8 (a) FDCB voltage, (b) input phase voltage, (c) output phase
voltage, (d) input phase current, (e) output phase current, (f)
harmonic characteristics of the input current and (g)
harmonic characteristics of the output current
122
4.9 SUMMARY
CHAPTER 5
5.1 INTRODUCTION
IA
VA
Ia Ib Ic
Va Vb Vc
VA
][ ]
V a S Aa SBa SCa SNa
[ ][
V b = S Ab SBb SCb
Vc S Ac SBc SCc SNc
V
SNb × B
VC
0
(5.1)
IA SAa SAb S Ac
[][
IB
IC
IN
S
= Ba
SCa
SNa
SBb
SCb
SNb
SBc
SCc
SNc
][]
Ia
× Ib
Ic
(5.2)
VDC+
IDC+
The FTC consists of three phase arms and one neutral arm.
Switching ON any of the two phase arms leads to the line voltage being
available at the FDCB and switching ON any one phase arm with the neutral
arm leads to the phase voltage being available at the FDCB. This results in
twelve active voltage vectors on the rectifier side. This decoupled
representation simplifies the control of the input current and the output
voltage in DTMC, as described in the next section.
127
[ ][ ][S S S S
SAb SBb SCb SNb = SIB SIb × CA CB CC CN (5.4)
SAc SBc SCc SNc SIC SIc
SCa SCb SCc SCn ]
The switching states for synthesizing the required currents and
voltages are described in the following subsections.
2π 4π
I α = I A + I B cos + I C cos (5.5)
3 3
2π 4π
I β = IB sin + I C sin
3 3
(5.6)
I0 = IA + IB + IC (5.7)
Vector IA IB IC I N = I0 I in ∠ I¿ V DC
SCa SCb SCc SCn
IZ
[ 11 00 0
0
0 0 1 0
][
0 0 1 0
0
0 ] 0 0
[ 00 00 1
1
0 0
0 ][ 0
0 0
0 0
1
1]
-1 Iβ
where, I in = √ I α2 + I β2 and Iin = tan .
Iα
ensure that the input current is sinusoidal, the reference space vector must lie
on the αβ plane requiring the neutral current to be zero on application of the
vector IPi. This is carried out by applying equally the adjacent I Pi vectors,
which lie on the upper and the lower halves of the αβ plane. This ensures that
the average neutral current is zero over a switching period. The example in
Table 5.2 explains the same.
Switching IA IB IC IN VDC
Applied vectors
time
Ts
I P6 0 - IDC 0 + I DC VNB
2
Ts
I P1 + I DC 0 0 - IDC VAN
2
1 1 1 1 1
Ts I VP1 = I P6 + I P1 + IDC - I DC 0 0 V
2 2 2 2 2 AB
IP4 IP2
IP6
IL3
IL4
IL2
IZ1, IZ2, IZ3, IZ4
IL1
IL5 IL6
IP3 IP1
IP5
IL3
2 1
IL4 IL2
IP3 IVP3 IP2
IVP4 IVP2
IP4
IP10
3
IVP5 IVP1
IP5 IP6
IVP6
IL5 IL1
4 5
IL6
Virtual Sharing
IA IB IC IN = I0 Iin Iin VDC
vectors vectors
1 1 √3 1
I VP1 [AN] I P6 , I P1 + IDC - I DC 0 0 I DC 3300 V
2 2 2 2 AB
1 1 √3 1
I VP2 [NC] I P1 , I P2 + IDC 0 - I DC 0 I DC 300 V
2 2 2 2 AC
1 1 √3 1
I VP3 [BN] I P2 , I P3 0 + IDC - I DC 0 I DC 90 0 V
2 2 2 2 BC
1 1 √3 1
I VP4 [NA] I P3 , I P4 - I DC + IDC 0 0 I DC 1500 V
2 2 2 2 BA
1 1 √3 1
I VP5 [CN] I P4 , I P5 - I DC 0 + IDC 0 I DC 2100 V
2 2 2 2 CA
1 1 √3 1
I VP6 [NB] I P5 , I P6 0 - I DC + IDC 0 I DC 2700 V
2 2 2 2 CB
Figure 5.4(a) shows the sector zero of the space vector diagram of
the FTC. Each sector consists of two active long vectors, two active virtual
short vectors and four zero vectors. To synthesize the required reference input
current and the FDCB voltage, the three nearest current vectors (Busquets-
Monge et al 2004) are selected, as shown in Figure 5.4(b), depending on the
modulation index mc of the FTC.
IL2 E
F
IVP2 IREF
R2
IZ R4
R5 G D
θC R3 H
A
R1
IVP1
B
IL1 C
E E
E
I1 I1
F F F I1
I3 R2 I3
F
I3 R4
I3
H D G H D
A I1 R5 H
G H
R3
I2
I2 R1
I2 I3
B B I2 B
B
I1 I2
C
C C
IL2 1 √3
cosec ( 600 + θc ) ; 00 <θc < 600
1/2 4
IVP2
IREF
1
IZ
√3/4 1/√3 √3/2 sec ( 600 - θc ) ; 300 <θc < 600
θC 2
IVP1 1
1/2 sec ( θc ) ; 0 0 < θc <30 0
IL1 1 2
Figure 5.5 Equations of lines used for identifying regions in the FTC
Conditions
1
1 R1 m c > sec ( θc ) ; 00 <θc < 300
2
133
1
2 R2 m c > sec ( 600 - θc ) ; 300 < θc <600
2
√3 1
3 R3 cosec ( 600 + θc ) < m c ≤ sec ( θc ) ; 00 <θc < 300
4 2
√3 1
4 R4 cosec (600 + θc ) < m c ≤ sec ( 600 - θc ) ; 30 0 < θc <60 0
4 2
√3
5 R5 mc≤ cosec ( 600 + θc ) ; 00 <θc < 600
4
x 1 x2 x 3 d 1 X
[ ][ ] [ ]
y1 y2 y3 d2 = Y
1 1 1 d3 1
(5.8)
IL2
Y IL2( cos600, sin600)
Y
IVP2
IREF
IVP2 (0.5 cos600, 0.5 sin600)
IZ IREF(mccosθc, mcsinθc)
θC X
θC
IVP1
X
IZ(0, 0) IVP1(0.5, 0) IL1(1, 0)
IL1
134
Duty Cycle
d1- I1 d2- I2 d3 - I3
2 4
R1 m sin ( θc ) 2 m c cos ( θc ) -1 2- m c sin ( 600 +θ c )
√3 c √3
2 4
R2 2 m c sin ( 300 + θc ) -1 m sin ( 600 - θc ) 2- m sin ( 600 +θ c )
√3 c √3 c
4 4
R3 m c sin ( 600 + θc ) -1 m c sin ( 600 - θc ) 2 - 4 m c cos ( θc )
√3 √3
4 4
R4 m c sin ( 600 + θc ) -1 2 - 4 m c sin ( 300 +θc ) m sin θc
√3 √3 c
4 4 4
R5 1- m sin ( 600 +θc ) m sin ( 600 - θc ) m sin θc
√3 c √3 c √3 c
where, θc is the angle of IREF within the sector and mc is the inverter zero
compensated converter modulation index discussed later and is given in
Equation (5.14).
where, θv is the angle of VOUT within a sector. The output voltage of the
inverter can be adjusted by any one of the two schemes: (i) changing the
FDCB voltage to the inverter, (ii) changing the modulation index m v of the
inverter. The second scheme is not used for reasons explained in the next
paragraph and hence mv=1. The FDCB voltage can be varied by changing the
modulation index m 'c of FTC as given in Equation (5.11)
135
Vb
V2 [110]
V3 [010] VMejπ/3Zero Vectors
VMej2π/3
2 V0 [000]
V7 [111] Vγ
3 1
V4 [011]
VMejπ Vγdγ Va Vγdγ
4 V1 [100] VREF
Vδdδ θv VOUT VMe-j0
0
θv Vδ
5 V0 d0
V5 [001] Vδdδ
VMe-j2π/3 V6 [101]
Vc VMe-jπ/3
Figure 5.7 (a) Space vectors of the FI and (b) sector and duty cycle
allocation
√3
m 'c = m DTMC (5.11)
2
dγ dδ
d 'γ = & d'δ =
d γ + dδ dγ + dδ
(5.12)
136
√3
mc= m DTMC ( d γ + dδ ) (5.14)
2
dγ dδ
Vγ Vδ
I3 I1 I2 I3 I3 I1 I2 I3
dδdV1
dδdV3 /2
dδdV2
dγdV3/2
dγdV3/2
dδdV3/2
dγdV1
dγdV2
Figure 5.8 (a) Switching pattern of the DTMC for the regions R1 and R2
137
dγ dδ
Vγ Vδ
I3 I2 I1 I2 I3 I3 I2 I1 I2 I3
dδdV3 /2
dγdV3 /2
dδdV2 /2
dδdV1
dγdV1
dγdV2 /2
dγdV2 /2
dδdV2 /2
dδdV3 /2
dγdV3 /2
Figure 5.8(b) Switching pattern of the DTMC for the regions R3, R4 and R5
400
C a lc u la te d F D C V
300
200
100
0
0.06 0.065 0.07 0.075 0.08 0.085 0.09 0.095 0.1
Time(s)
400
Calculated FDCV (V)
300
200
100
0
0.06 0.065 0.07 0.075 0.08 0.085 0.09 0.095
Time(s)
Figure 5.9 (b) Calculated FDCB voltage with an unbalanced input voltage
√3
mc= m DTMC ( d γ + dδ ) × VDC_Min / V DCF (5.16)
2
the device during its transition (‘ON’ to ‘OFF’ or vice-versa) states is called
the switching loss. Conduction loss is the product of the voltage drop across
the device and the current through the device, when it is in the ‘ON’ state.
Switching loss is proportional to the product of blocking voltage and
conduction current at the instant of switching; and if this is significant, it is
termed as hard switching loss (Bierhoff and Fuchs 2004). If the switching
occurs when either the current through the device or the voltage across the
device is nearly zero, the commutation is referred to as ‘soft switching’ and
the switching loss in the device is negligible. For an IGBT, there are two
types of losses during hard switching: Ton_losses and Toff_losses, associated with the
device turn-ON and turn-OFF process respectively. For a diode, the switching
loss is caused by reverse recovery mechanism that occurs only during the
diode turn-OFF. Hence, the turn-ON loss for a diode is not considered.
5.6.1 Conduction Loss Modeling for the CMC and the DTMC
From Equation (5.3), it can be seen that in each phase only one
switch conducts at any given time. Hence, there is always only one IGBT that
conducts and only one diode that conducts at an output phase of the CMC and
the DTMC. Equations (5.17) and (5.18) give the conduction loss and the
conduction energy of one output phase in each switching cycle
CLosses ( v d , i L ) = v d ( i L ) × i L + vd
IGBT DIODE
( iL ) × i L (5.17)
Ts
Ec ( v d , i L )=∫ CLosses ( v d , i L ) dt (5.18)
0
v d ( i L ) =x + y × i Lz
IGBT
(5.19)
140
vd DIODE
( iL ) =m + n × i Lk (5.20)
where, x, y, z, m, n and k are constants that are obtained from the curve fitting
equation of Vce-Ic characteristics given in the datasheet of the device used.
Then the average conduction loss over an interval T, for the CMC and the
DTMC, is give by Equation (5.21).
T
1
CL_Avg = ∫ CLosses ( t ) dt (5.21)
T 0
5.6.2 Switching Loss Modeling for the CMC and the DTMC
where, VR, iR and EswR are respectively the voltage, current and switching
energy of the device at the rated VR and iR. From Figure 1.4, and the four step
commutation procedure, discussed in chapter 1, it can be seen that when
commutation happens between the bidirectional switch S1 to switch S2 under
the condition of Vin >0 and Iout >0, commutation losses do not occur for
switches S1-, S2+ and S2-. This is because the switches S1- and S2- do not block
any voltage and the switch S 2+ does not conduct current. This creates only a
turn OFF loss for the switch S1+. Similarly, S2 to S1 transition creates a turn
ON loss for the switch S1+ and a turn OFF loss for the diode D2-. Table 5.6
summarizes the switching energy losses for commutation between S 1 and S2
evaluated for all conditions of input voltages and output currents.
Switch S1 S2 S2 S1 S1 S2 S2 S1
141
transition
Iout + Iout -
Vin + Eoff Eon + E rr_D Eon + E rr_D Eoff
Vin - Eon + E rr_D Eoff Eoff Eon + E rr_D
(5.23)
From the Tdelay-Ic characteristics of the datasheet, Ton, Toff and Tr are
identified. Equation (5.24) gives the switching power loss.
Ton T off T r
(5.24)
output phases in a switching cycle Ts. Four of these commutation events occur
in an output phase and two commutation events each occur in the other two
output phases.
From Figure 5.10, it can be seen that the total switching energy of
the CMC over a sampling time Ts is given by Equation (5.25)
where, K= (Eon+ Eoff + Err_D) / (VR × iR). For other voltage and current sectors,
Equation (5.25) can be generalized as Equation (5.26)
Esw = K ( x i a + y i b + z i c ) (5.26)
where, x, y and z take any of the values |v AB|,| vBC |,| vAC |, (|vAB|+| vBC|) or (|vAB|
+| vAC|) depending on the sectors of the current and the voltage.
Esw 1 = K . (|v AN|. i a +(|v BC|+|v CN|+|v BN|). i b + (|v AB|+|v AC|+2|v AN|+|v BN|+|v CN|) . i c )
R
(5.27)
Region 2
Esw = K . (|v AN| . i a +(|v BC|+ 2|v CN|). i b+ (|v AB|+|v AC|+ 2|v AN|+2|v CN|) . i c )
R2
(5.28)
Region 3
Esw = K . (3|v AN|. i a +(|v BC|+|vCN|+|v BN|). i b+ (|v AB|+ 3|v AN|+|v BN|+ 2|v CN|) . i c )
R3
(5.29)
Region 4
144
Esw 4 = K . (3|v AN|. i a +(|v CN|+|v BN|). i b+ (|v AC|+ 3|v AN|+|v BN|+ 2|v CN|) . i c )
R
(5.30)
Region 5
Esw = K . (2|v AN|. i a +(|v BN|+2|v CN|). i b + (|v BN|+2|v AN|+2|v CN|) . i c )
R5
(5.31)
The switching and conduction losses for the DTMC were derived
and compared with those for the CMC. A complete loss model was developed
using the Simulink blockset in MATLAB. Through simulations, switching
energy losses for different regions for different sectors of the current and the
voltage are calculated using the above procedure and results obtained are
discussed and presented in the next section.
5.7 SIMULATION
Quantity Value
R-L Load R = 20Ω , L = 21mH
Input phase voltage 100 V
Input voltage frequency 50 Hz
Input filter L = 2 mH, C = 35 µF, Rd = 15 Ω
Output Voltage frequency 25 Hz
Switching frequency 6 kHz
Modulation Index 0.72, 0.5, 0.25
145
√3 √3
For modulation indices between and , the output voltage
4 2
switches between the active long vectors and the active short vectors but for
lower modulation indices, the output voltage switches between the active
short vectors and the zero vectors. Figure 5.11 shows the output phase
voltages, output line voltages, input currents and output currents for the
voltage transfer ratio that is changed from 0.72 to 0.5 at 0.4 s and 0.5 to 0.25
at 0.5 s. The harmonic content of the input and the output currents increase
with decrease in the voltage transfer ratio. In the CMC, the peak of the output
voltage is √ 3 times the input voltage for all values of modulation indices.
However, in the DTMC, the peak of the output voltage is √ 3 times the input
√3
voltage for the modulation indices greater than while the peak of the
4
√3
output voltage is times the input voltage for modulation indices lesser
2
√3
than . This leads to lower switching stress on the power devices in the case
4
of the DTMC.
200
100
Van (V)
-100
-200
0.3 0.35 0.4 0.45 0.5 0.55 0.6
Time(s)
(a)
146
200
100
Vab (V)
-100
-200
0.3 0.35 0.4 0.45 0.5 0.55 0.6
Time (s)
(b)
2
IABC (A)
-2
-4
0.3 0.35 0.4 0.45 0.5 0.55 0.6
Time (s)
(c)
2
Iabc (A)
-2
-4
0.3 0.35 0.4 0.45 0.5 0.55 0.6
Time (s)
(d)
100
50
VABC ( V )
-50
-100
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time (s)
(a)
0
-50
-100
-150
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time (s)
(b)
148
IABC (A) 2
-2
-4
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time (s)
(c)
4
2
Iabc ( A)
-2
-4
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time (S)
(d)
0.9
Modulation Index (c)M
0.7
0.5
0.3
0
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time(s)
(e)
Figure 5.12 Performance of the DTMC with an unbalanced supply
(a) output phase voltage, (b) output line voltage, (c) input
phase current and (d) output phase current and
(e) modulation index
Figures 5.13 to 5.15 present a quantitative comparison between the
CMC and the DTMC. Simulation was carried out for the CMC and the
DTMC for a Modulation Index (MI) of 0.866 and the THD of the output
voltage are shown in Figures 5.13(a) and 5.13(b). At the maximum
modulation index of 0.866, the THD for the DTMC reduces by 10% when
compared to the CMC. The THD content of the output voltage for the CMC
and the DTMC, for all modulation indices, is presented in Figure 5.13(c). It
149
can be seen that the DTMC has a better (lower) THD than the CMC for all
modulation indices. In the DTMC, for modulation indices varying from 0.866
to 0.45, the current vector lies in any of the regions of R 1, R2, R3, or R4 and the
THD is almost constant as zero vectors are not selected. When the current
vector is in the region R5 (MI < 0.4), the THD of the DTMC rises linearly
with modulation index, as in the case of the CMC, because of the use of the
zero vectors.
25
20
15
10
0
0 1000 2000 3000 4000 5000 6000 7000
Frequency (Hz)
(a)
Fundamental (25Hz) = 345.1 , THD= 53.20%
Mag (% of Fundamental)
20
15
10
0
0 1000 2000 3000 4000 5000 6000 7000
Frequency (Hz)
(b)
Figure 5.13 (Continued)
150
(c)
Figure 5.13 Output voltage THD for MI of 0.866 (a) CMC, (b) DTMC,
(c) output voltage THD for the CMC and the DTMC with 6
kHz switching frequency for different MI
At very low modulation index, the THD for the DTMC reduces by
approximately 58% as compared to the CMC due to the use of phase vectors.
From Figure 5.14(a), it can be observed that the conduction losses are always
greater than the switching losses in the CMC, for different values of MIs.
Figure 5.14(b) shows that as the input power factor decreases, the output
power of the converter also decreases.
(a)
Figure 5.14 (Continued)
151
(b)
Figure 5.14 (a) Losses vs. MI and (b) output vs. MI (different IPF)
The conduction losses for the DTMC and the CMC are the same
under all operating conditions. However, the switching losses for the DTMC
are higher than the switching losses of the CMC for all MIs, since the
switching events are more in the DTMC than in the CMC. With a double side
banded SVM, the DTMC exhibits higher switching losses for all values of
MIs. Nevertheless, for the single sided SWM, the DTMC exhibits higher
switching losses for lower values of MIs and lower switching losses for
higher values of MIs. This is because the switching events of the regions R 3,
R4 and R5 are very high compared to the switching events of the regions R 1
and R2, as shown in Figure 5.15.
152
Figure 5.15 Conduction and switching losses for the DTMC and the
CMC for different values of MI
(a)
(b)
(c)
(d)
Figure 5.17 Hardware output for balanced input condition (a) output
phase voltage, (b) output line voltage, (c) output phase
current and (d) input phase current
155
5.7 SUMMARY
In this chapter, the space vector PWM technique for the direct
three-level matrix converter has been proposed for synthesizing balanced
sinusoidal three-level output voltages from balanced and unbalanced
non-sinusoidal input voltages. In addition, conduction losses and switching
losses were modelled for the DTMC and a comparative study of the same for
the CMC and the DTMC has been carried out.
CHAPTER 6
6.1 INTRODUCTION
a d ψ as
a a
V =R i +
s s s + j ω a ψs (6.1)
dt
a d ψar a
0 = Rr ir + + j ( ωa - ωr ) ψ s (6.2)
dt
ψas = Ls i as + Lm i ar (6.3)
ψar = Lr i ar + Lm i as (6.4)
3
Te = P ψ as × i as (6.5)
2
d ωm J d ωr
T e - TL = J = (6.6)
dt P dt
where, vs, ψs, ψr, is and ir are the stator voltage, stator flux, rotor flux, stator
current and rotor current vectors respectively; ω r and ωm are the rotor speed in
rad/s and the mechanical speed in rpm respectively; L s, Lr and Lm are the
stator, rotor and magnetizing inductances respectively; T e, TL, J and P are the
159
electromagnetic torque, load torque, system inertia and the number of pole
pairs respectively. These space phasor quantities are expressed in the arbitrary
reference frame, which rotates at an arbitrary speed. The superscript ‘a’
denotes that the quantities are in the arbitrary reference frame. In the
stationary reference frame, the stator voltage vector given in Equation (6.1) is
replaced by Equation (6.7), where the superscript ‘s’ denotes that the
quantities are in the stationary reference frame.
s s d ψss
V s = Rs i s + (6.7)
dt
Assuming that over a small period of time, the voltage drop across
the stator resistance can be neglected, Equation (6.7) can be rewritten as
Equation (6.8).
Δ ψs
Vs = (6.8)
Δt
From Equation (6.5), the relationship between the stator flux, the
rotor flux and the electromagnetic torque can be expressed by Equation (6.9).
3
Te= P |ψ s| |ψ r|sin θsr (6.9)
2
The fast variation of the stator flux compared to that of the rotor
flux causes the stator flux to lead the rotor flux. Therefore, over a small period
of time, the rotor flux is assumed to have a slower rotation compared to the
stator flux. For this reason, when the appropriate space vector (voltage vector)
is applied to a power converter, the stator flux will change quickly. Hence, the
160
angle between the flux vectors θsr will increase, which will cause an increase
in the electromagnetic torque Te . In other words, the variation of the stator
voltage will affect both ψs and θsr . This principle is used in the DTC to achieve
the desired torque response and correct the flux trajectory of the IM.
With the input voltages, the line currents and the present switch
position as the inputs, the estimator model (Bose 2004) calculates the actual
flux, torque and speed of the motor. The errors in the flux and the torque are
fed to the two-level flux comparator and the three-level torque comparator
respectively, as shown in Figures 6.1(a) and 6.1(b). In the classical DTC, to
reduce the switching frequency and the torque ripple, the three-level
hysteresis band (Beerten et al 2009) is used.
+1 +1
0 ΔΨs 0 ΔTe
0 0
-1 Ψs_ref -1
Ψs - ΔΨs Ψs + ΔΨs Te – ΔTe TL Te +ΔTe
(a) (b)
respectively the torque, based on the sector (position) in which the stator flux
is present. Figure 6.2(c) shows that when the error in the electromagnetic
torque touches the zero level of the hysteresis band, the zero voltage vectors
are applied to maintain the torque constant. Practically the torque does not
remain constant due to the losses in the machine, but still a zero voltage
vector is applied under such condition in order to reduce the switching
frequency and to minimize the torque ripple.
Vb
γ
V3 [010] V2 [110]
3 1
Va
V4 [011] V1 [100]
4 6
5
V5 [001] V6 [101] V0 [000]
V7 [111]
Vc
qs
V4 . Δt4
E
D V3. Δt3
V2 . Δt2
B
C
V3. Δt1
S(3) S(2) A
π/3
S(1)
S(4) ds
S(6)
S(5)
2HBΨ
Te + ΔTe
TL= 3 Nm
Te – ΔTe
TL =0
Table 6.1 Optimal switching table of the VSI for the DTC
Figure 6.3 shows the space vectors for the MLI. Table 6.2 gives the
extended OST of the MLI for the DTMC. In order to reduce the torque ripple,
the short vectors of the MLI are also used in addition to the large vectors.
Thus, the DTC scheme is modified to have a new torque hysteresis
comparator that will provide five different levels instead of three levels, to
distinguish between the small and the large positive and negative torque
errors, as shown in Figure 6.4.
ΔΨ3L
Vb
ΔΨ2L
V2 L
ΔΨ3S
V3 L ΔΨ4L ΔΨ1S ΔΨ1L
V3 S V2 S ΔΨ5S
2 ΔΨ6L
3 1 ΔΨ5L
Va
V4L V4S 4 6 V1S V1L
5
V5 S V6 S
V5 L V6 L
Vc
Figure 6.3 MLI space vector and the corresponding variation in the
flux
164
HT
2
1
-LEr -SEr ET
SEr LEr
-1
-2
Table 6.2 Extended optimal switching table of the MLI for the DTC
Sector of Ψ
HΨ=-1 HΨ=1
HT=-2 HT=-1 HT=0 HT=1 HT=2 HT=-2 HT=-1 HT=0 HT=1 HT=2
As in the VSI, the CMC DTC scheme uses the same reference
quantities such as the flux and the torque that are controlled by selecting the
appropriate output voltage vectors. In addition, the control of the input current
that is essential in the CMC is initiated by selecting the appropriate input
165
current vector. Hence, the choice of the switching configuration for the OST
is made on a different basis as described next (Matteini 2001).
Once the classical DTC scheme of the inverter has selected the
optimum vector to be applied to the machine, it is a matter of determining the
corresponding matrix converter switching configuration. For example, if the
VSI output vector V1 has been chosen, from Table 6.1, Figure 6.2(a) and
Figure 6.5(a), it can be seen that the CMC can generate the same vector by
means of the switching configurations ±1, ±2, and ±3. Keeping in mind the
input current control we can see that not all of them can be usefully employed
to provide an equivalent of the vector V1.
Vb VB
±4, ±5, ±6 ±2, ±5, ±8
2 2
3
3 1 ±1, ±2, ±3 1
4 VA
4 Va
6
5 6
5 ±1, ±4, ±7
±3, ±6, ±9
±7, ±8, ±9 VC
Vc (a) (b)
Figure 6.5 (a) CMC output voltage vectors and (b) input current vectors
VSI output vector chosen by the classical DTC scheme without affecting the
current direction. This redundancy can be used to control the input power
factor in addition to the control of the stator flux and the electromagnetic
torque.
±2, ±5, ±8
+1
0 Sin (α)
0 -3
-1 ei
Ψ=0
Sin (ψ - θ) Sin ψ Sin (ψ + θ)
i
+1
±3, ±6, ±9 ±1, ±4, ±7
(a) (b)
Figure 6.6 (a) Sine-hysteresis comparator and (b) leading and lagging
vectors for the input current control
167
Table 6.3 Optimal switching table of the CMC for the DTC
±4, ±5, ±6
2
±1, ±2, ±3
4 6
±2, ±5, ±8
3 2
4 1
±10, ±13, ±16
5 6
The 18 long vectors are numbered as ±1, ±2, ±3, ±4, ±5, ±6, ±7, ±8
and ±9. Similarly, the 18 short vectors are numbered as ±10, ±11, ±12, ±13,
±14,±15, ±16, ±17 and ±18.
171
The classical DTC scheme of the MLI is used as the basis for the
DTMC DTC scheme. Once the optimum vector to be applied to the IM is
selected from Table 6.2, it is only a matter of determining the corresponding
DTMC switching configuration. If the selected MLI vector is a long
vector VL, then the equivalent long vector of the DTMC is selected from
Table 6.3. Table 6.4 shows that the long vectors of the DTMC do not
contribute to the neutral current and this allows the use of the same procedure
of the CMC for selecting the equivalent DTMC long vector.
One of the problems in the use of the DTMC is the input filter
capacitor voltage imbalance. To understand this issue, we have to study the
effect of each type of short voltage vectors on the deviation of neutral point
voltage VN and current IN. Let us analyze the deviation of V N due to the two
short vectors, +10(ann) and -11(nbb). As illustrated in Figure 6.8, the sign of
the neutral point current (IN) determines the sign of the neutral point voltage
deviation (Δv) given by Equation (6.10).
172
Δv = V cA + V cB + V cC (6.10)
A A
B
Ia B Ia
C C a
a
VcB
VAN VNB
VcA VcC VcA VcC
VcB
IN IN b c
b c
-Ia -Ia
(a) (b)
Figure 6.8 Neutral current (IN) for (a) space vector +10 (ann) and (b)
space vector -11 (nbb)
+1
0 Δv
-1
-V0 0 +V0
It can be concluded that once the classical DTC scheme of the MLI
selects the voltage vectors to be applied to the IM, it is only a matter of
173
Table 6.5 Optimal switching table of the DTMC short vectors Vs for the
DTC
1 2 3 4 5 6
Bψ
+1 -1 +1 -1 +1 -1 +1 -1 +1 -1 +1 -1
BV
V1S +1 +10 +10 -12 -12 +11 +11 -10 -10 +12 +12 -11 -11
-1 -12 -11 +11 +10 -10 -12 +12 +11 -11 -10 +10 +12
S
V2 +1 -16 -16 +18 +18 -17 -17 +16 +16 -18 -18 +17 +17
-1 +18 +17 -17 -16 +16 +18 -18 -17 +17 +16 -16 -18
S
V3 +1 +13 +13 -15 -15 +14 +14 -13 -13 +15 +15 -14 -14
-1 -15 -14 +14 +13 -13 -15 +15 +14 -14 -13 +13 +15
S
V4 +1 -10 -10 +12 +12 -11 -11 +10 +10 -12 -12 +11 +11
-1 +12 +11 -11 -10 +10 +12 -12 -11 +11 +10 -10 -12
S
V5 +1 +16 +16 -18 -18 +17 +17 -16 -16 +18 +18 -17 -17
-1 -18 -17 +17 +16 -16 -18 +18 +17 -17 -16 +16 +18
S
V6 +1 -13 -13 +15 +15 -14 -14 +13 +13 -15 -15 +14 +14
-1 +15 +14 -14 -13 +13 +15 -15 -14 +14 +13 -13 -15
Figure 6.10 shows the block diagram of the proposed DTC scheme
for the DTMC. The estimators shown in Figure 6.10 require the knowledge of
the input and the output voltages and currents. However, only the input
voltages and the output currents are measured in each cycle period, because
the other quantities can be calculated from the actual switching configuration
of the DTMC.
174
Vin Vin
Vin
Vin_Sector
Vin_Sector
DTMC POWER
CONVERTER
CT VL
Tref T
CΨ
OR
T Cφ Vin_Sector
Table 6.3
φref
VS
φest
T CV
ΔV ref
Test φest Sin Ψest ΔV
ΔV
Torque , Flux, sinΨ Estimators & {ΔV}-Generator
Induction Motor
Figure 6.10 Block diagram of the proposed DTC scheme for the DTMC
3
Te= L (i i - i i ) (6.11)
2 m ds qr qs dr
Since the frequency of the rotor current is very lower than the stator
current, it can be concluded that the dynamics of the torque is directly
affected by the dynamics of the stator current and vice-versa. Hence, the
harmonic content in the stator current can be taken as a measure of the ripple
content in the torque of the IM.
175
6.6 SIMULATION
The analysis has shown that the proposed control scheme can
provide good performance for the IM at unity input power factor. The input
line current can be significantly distorted if the input power factor is not
closer to unity and the line currents also get distorted even if the sampling
frequency is not sufficiently high.
150
Sp eed (rad /sec)
100
50
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time(s)
(a)
10
Torq ue (N -m ) 5
-5
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
(b)
10
5
Ia (A)
-5
-10
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
(c)
40
VA (V) , IA (A)
20
-20
-40
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
40 40
20 20
VA (V) , IA (A)
VA (V) , IA (A)
0 0
-20 -20
-40 -40
0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9 1.3 1.31 1.32 1.33 1.34 1.35 1.36 1.37 1.38 1.39 1.4
Time (s) Time (s)
(d)
-2
-4
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s)
4
Voltage Deviation ( V)
-2
-4
0.5 0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6
Time (s)
(e)
Figure 6.11 (a) Speed of the IM, (b) torque of the IM (c) input current of
the IM, (d) scaled input voltage (230/6) and input current of
the DTMC, (e) filter capacitor voltage deviation
(a) (b)
Figure 6.12 (a) Variation of flux (a) stator and (b) rotor
178
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Frequency (Hz)
(a)
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Frequency (Hz)
(b)
8
Torque (N -m )
0
0.95 1 1.05 1.1 1.15 1.2 1.25
Times (s)
(a)
Figure 6.15 (Continued)
180
10
T o rq u e (N -m )
8
0
0.95 1 1.05 1.1 1.15 1.2 1.25
Time (s)
(b)
Figure 6.15 Response to a step torque command (a) CMC-DTC
and (b) DTMC-DTC
6.7 SUMMARY
CHAPTER 7
7.1 CONCLUSION
during the application of the short vectors and minimizes these fluctuations.
Simulation results portray the improved performance of the DTC scheme for
the DTMC.
APPENDIX 1
Figure A1.1 Current and voltage sensors used for the DTMC topology
Figure A1.2 Analog data acquisition board used for the DIDC PWM
technique
186
Figure A1.3 Matrix converter power module used for the DIDC – PWM
technique
Figure A1.4 Direct three level matrix converter (DTMC) power module
187
APPENDIX 2
With the same rules the modified optimum indirect SVM for the
DTMC is formulated which ensured that the switching transitions are
minimum.
where, γ and δ indicates the duty cycles of the active line vectors and Q 1and
Q2 indicates the duty cycles of the active phase vectors.
190
βγ αγ αδ βδ Θ βδ αδ αγ βγ
abb aba aca acc acc aca aba abb
Tβγ/2 Tαγ/2 Tαδ/2 Tβδ/2 Tθ Tβδ/2 Tαδ /2 Tαγ/2 Tβγ/2
V1 – IL1 V6 - IL1 V6-IL2 V1-IL2 V1-IL1 V6-IL2 V6-IL1 V1-IL1
SAa SAc
SAb
SBb
SBc
SBa
SCb
SCa SCc
The energy lost in switching in input current sector -1, region -1and
voltage sector-1 is given by the Equation (A2.1)
Esw 1 = K . (|v AN|. i a +(|v BC|+|v CN|+|v BN|). i b + (|v AB|+|v AC|+2|v AN|+|v BN|+|v CN|) . i c )
R
(A2.1)
Esw = K . (3|v AN|. i a +(|v BC|+|vCN|+|v BN|). i b+ (|v AB|+ 3|v AN|+|v BN|+ 2|v CN|) . i c )
R3
(A2.2)
The energy lost in switching in input current sector -1, region -5 and
voltage sector-1 is given by the Equation (A2.3).
Esw = K . (2|v AN|. i a +(|v BN|+2|v CN|). i b + (|v BN|+2|v AN|+2|v CN|) . i c )
R5
(A2.3)
194
APPENDIX 3
REFERENCES
3. Apap, M., Clare, J.C., Wheeler, P.W. and Bradley, K.J. “Analysis and
comparison of ac-ac matrix converter control strategies”, Proceedings
of 34th IEEE Power Electronics Specialists Conference, Acapulco,
Mexico, Vol. 3, pp. 1287-1292, 2003.
16. Bowes, S.R. and Lai, Y.S. “The relationship between space-vector
modulation and regular-sampled PWM”, IEEE Transactions on Power
Electronics, Vol. 14, No. 5, pp. 670-679, 1997.
19. Cardenas, R., Pena, R., Wheeler, P., Clare, J. and Asher, G. “Control
of the reactive power supplied by a WECS based on an induction
generator fed by a matrix converter”, IEEE Transactions on Industrial
Electronics, Vol. 56, No. 2, pp. 429-438, 2009.
20. Casadei, D., Grandi, G., Serra, G. and Tani, A. “Switching strategies
in direct torque control of induction machine”, Proceedings of
International Conference on Electrical Machines, Paris, France, Vol .2,
pp. 204-209, 1994.
21. Casadei, D., Serra, G. and Tani, A. “Reduction of the input current
harmonic content in matrix converters under input/output unbalance”,
IEEE Transactions on Industrial Electronics, Vol. 45, pp. 401-411,
1998.
23. Casadei, D., Serra, G. and Tani, A. “The use of matrix converters in
direct torque control of induction machines”, IEEE Transactions on
Industrial Electronics, Vol. 48, No. 6, pp. 1057-1064, 2001.
24. Casadei, D., Serra, G., Tani, A. and Zarri, L. “Matrix Converter
modulation strategies: A new general approach based on space - vector
representation of the switch state”, IEEE Transactions on Industrial
Electronics , Vol. 49, pp. 370-381, 2002.
27. Cha, H. J. and Prasad, N.E. “An Approach to reduce common mode
voltage in matrix converter”, IEEE Transitions on Industry
Applications, Vol. 39, No. 4, pp. 1151 -1159, 2003.
36. Fa-Hai Li, Yao-Hua Li, Xiang-Heng Wang, and Jing-De Gao,
“Modelling and simulation of synchronous motor fed by
cycloconverter”, Proceedings of 25th IEEE Power Electronics
Specialists Conference, Taipei, Taiwan, Vol. 2, pp. 830-834, 1994.
45. Holmes, D.G. “The significance of zero space vector placement for
carrier-based PWM schemes”, IEEE Transactions on Industrial
Electronics, Vol. 32, No. 5, pp.1122-1129, 1996.
54. Itoh, J., Sato, I., Odaka, A., Ohguchi, H., Kodachi, H. and Eguchi, N.
“A novel approach to practical matrix converter motor drive system
with reverse blocking IGBT”, IEEE Transactions on Power
Electronics, Vol. 20, No. 6, pp. 1356-1362, 2005.
58. Kim, J. and Sul, S. “New control scheme for AC-DC-AC converter
without DC-link electrolytic capacitor”, Proceedings of 24 th IEEE
Power Electronics Specialist Conference, Seattle, USA, pp. 300-306,
1993.
204
59. Kim, Y. and Sul, S. “Novel ride through system for adjustable speed
drives using common mode voltage”, IEEE Transactions on Industry
Applications, Vol. 37, No. 5, pp. 1373-1382, 2001.
62. Klumpner, C., Nielsen, P., Boldea, I. and Blaabjerg, F. “A new matrix
converter motor (MCM) for industry applications”, IEEE Transactions
on Industrial Electronics, Vol. 49, No. 2, pp. 325-335, 2002.
63. Kolar, J., Baumann, M., Schafmeister, F. and Ertl, H. “Novel three-
phase AC-DC-AC sparse matrix converter”, Proceedings of 17 th IEEE
Applied Power Electronics Conference, Dallas, USA, Vol. 2, pp. 777-
701, 2002.
66. Lie Xu and Cartwright, P. “Direct active and reactive power control of
DFIG for wind energy generation”, IEEE Transactions on Energy
Conversion, Vol. 21, No. 3, pp. 750-758, 2006.
68. Lopez, O., Alvarez, J., Gandoy, J. D. and Freijedo, F.D. “Multilevel
multiphase space vector pwm algorithm”, IEEE Transactions on
Industrial Electronics, Vol. 55, No. 5, pp. 1933-1942, 2008.
205
69. Lozano, J.M. and Ramirez, J.M. “Matrix converter performance under
unbalanced input-voltage”, Proceedings of 40th North American
Conference on Power Energy and Industry Applications, Alberta,
Canada, pp. 1-6, 2008.
71. Mahlein, J., Simon, O. and Braun, M. “A matrix converter with space
vector control enabling over modulation”, Proceedings of European
Conference on Power Electronics and Drives, Lausanne, Switzerland,
conference CD, 1999.
72. Mahlein, J., Igney, J., Weigold, J., Braun, M. and Simon, O. “Matrix
converter commutation strategies with and without explicit input sign
measurement”, IEEE Transactions on Industrial Electronics, Vol. 49,
No. 2, pp. 407-414, 2002.
75. Mekhilef, S. and Kadir, M.N.A. “Novel vector control method for
three-stage hybrid cascaded multilevel inverter”, IEEE Transactions on
Industrial Electronics, Vol. 58, No. 4, pp. 1339-1349, 2011.
81. Neft, C.L. and Schauder, C.D. “Theory and design of a 30-HP matrix
converter”, IEEE Transactions on Industry Applications, Vol. 28,
No.3, pp. 546-551, 1992.
85. Ogasawara, S., Ayano, H. and Akagi, H. “An active circuit for
cancellation of common-mode voltage generated by a PWM inverter”,
IEEE Transactions on Power Electronics, Vol. 13, No. 5, pp. 835-841,
1998.
87. Oyama, J., Higuchi, T., Yamada, E., Koga T. and Lipo, T. “New
Control Strategy for Matrix converter”, Proceedings of 20 th IEEE
Power Electronics Specialist Conference, Milwaukee, USA, pp. 360-
367, 1989.
88. Podlesak, T., Katsis, D., Wheeler, P., Clare, J., Empringham, L. and
Bland, M. “A 150-kVA vector-controlled matrix converter induction
motor drive”, IEEE Transactions on Industry Applications, Vol. 41,
No. 3, pp. 841-847, 2005.
91. Rojas, R., Ohnishi, T. and Suzuki, T. “Simple structure and control
method for a neutral-point-clamped pwm inverter”, Proceedings of
Power Conversion Conference, Yokohama, Japan, pp. 26-31, 1993.
99. Videt, A., Le Moigne, P., Idir, N., Baudesson, P. and Ecrabey, J. “A
New Carrier-Based PWM for the Reduction of Common Mode
Currents Applied to Neutral-Point-Clamped Inverters”, Proceedings of
22nd IEEE Applied Power Electronics Conference, pp. 1224-1230,
2007.
208
102. Wang, H., Zhao, R., Cheng, F., Yang, H. “Six-phase induction
machine driven by the matrix converter”, Proceedings of IEEE
Electrical Machines and Systems Conference, Beijing, China, pp. 1-5,
2011.
103. Wei, L. and Lipo, T.A. “A Novel matrix converter with simple
commutation”, Proceedings of 36th IEEE Industry Applications Society
Conference, Chicago, USA, Vol. 3, pp. 1749-1754, 2001.
104. Wei, L., Lipo, T.A. and Chan, H. “Matrix converter topologies with
reduced number of switches”, Proceedings of 33rd IEEE Power
Electronic Specialists Conference, Queensland, Australia, Vol. 1, pp.
57-63, 2002.
107. Wheeler, P., Clare, J., Empringham, L., Bland, M. and Apap, M. “Gate
drive level intelligence and current sensing for matrix converter current
commutation”, IEEE Transactions on Industrial Electronics, Vol. 49,
No. 2, pp. 382-389, 2002b.
108. Wheeler, P. W., Empringham, L., Apap, M., De Lillo, L., Clare, J. C.,
Bradley, K. J., Whitley, C. and Towers, G. “A matrix converter based
permanent magnet motor drive for an aircraft actuation system”,
Proceedings of IEEE Electrical Machines and Drives Conference,
Wisconsin Madison, USA, Vol. 2, pp. 1295-1300, 2003.
112. Yoon, Y.D. and Sul, S.K. “Carrier - Based modulation technique for
matrix converter,” IEEE Transactions on Power Electronics, Vol. 21,
No. 6, pp. 1691-1703, 2006.
LIST OF PUBLICATIONS
International Journals
International Conferences
CURRILUM VITAE
and Electronics Engineering from the Madras University, Chennai in the year