A Refined Space Vector PWM Signal Generation for Multilevel Inverters

ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 2011

A Refined Space Vector PWM Signal Generation for
Multilevel Inverters
G.Sambasiva Rao, Dr.K.Chandra Sekhar
Dept. of Electrical Engg, R.V.R & J.C College of Engg, Chowdavaram, Guntur -522 019(India)
E-Mail:sambasiva.gudapati@gmail.com, cskoritala@gmail.com

Abstract— A refined space vector modulation scheme for inverter. The inverter leg switching times are calculated
multilevel inverters, using only the instantaneous sampled directly from the sampled amplitudes of the sinusoidal
reference signals is presented in this paper. The proposed reference signals with considerable reduction in the
space vector pulse width modulation technique does not require computation time [8]. The SPWM scheme, when applied to
the sector information and look-up tables to select the multilevel inverters, uses a number of level-shifted carrier
appropriate switching vectors. The inverter leg switching times
signals to compare with the sinusoidal reference signals [9,
are directly obtained from the instantaneous sampled
reference signal amplitudes and centers the switching times
19]. The SVPWM for multilevel inverters [10, 11] involves
for the middle space vectors in a sampling time interval, as in mapping of the outer sectors to an inner subhexagon sector,
the case of conventional space vector pulse width modulation. to determine the switching time interval, for various space
The simulation results are presented to a five-level inverter vectors. Then the switching space vectors corresponding to
system for dual-fed induction motor drive. The dual-fed the actual sector are switched, for the time durations calculated
structure is realized by opening the neutral-point of the from the mapped inner sectors. It is obvious that such a
conventional squirrel cage induction motor. The five-level scheme, in multilevel inverters, will be very complex, as a
inversion is obtained by feeding the dual-fed induction motor large number of sectors and inverter vectors are involved.
with four-level inverter from one end and two-level inverter
This will also considerably increase the computation time. A
from the other end.
Index Terms— dual-fed induction motor, space vector PWM,
modulation scheme is presented in [12], where a fixed common
sampled sinusoidal reference signals, triangular carrier mode voltage is added to the reference signal throughout
signals, middle space vectors. the modulation range. It has been shown [13] that this common
mode addition will not result in a SVPWM-like performance,
I. INTRODUCTION as it will not centre the middle space vectors in a sampling
interval. The common mode voltage to be added in the
The two most widely used pulse width modulation (PWM) reference phase voltages, to achieve SVPWM-like
schemes for multilevel inverters are the carrier-based sine- performance, is a function of the modulation index for
triangle PWM (SPWM) scheme and the space vector PWM multilevel inverters [13]. A SVPWM scheme based on the
(SVPWM) scheme. These modulation schemes have been above principle has been presented [14], where the switching
extensively studied and compared for the performance time for the inverter legs is directly determined from sampled
parameters with two level inverters [1, 2]. The SPWM schemes sinusoidal reference signal amplitudes. This technique
are more flexible and simpler to implement, but the maximum reduces the computation time considerably more than the
peak of the fundamental component in the output voltage is conventional SVPWM techniques do, but it involves region
limited to 50% of the DC link voltage [2]. In SVPWM schemes, identifications based on modulation indices. While this
a reference space vector is sampled at regular intervals to SVPWM scheme works well for a three-level PWM
determine the inverter switching vectors and their time generation, it cannot be extended to multilevel inverters of
durations, in a sampling time interval. The SVPWM scheme levels higher than three, as the region identification becomes
gives a more fundamental voltage and better harmonic more complicated. A carrier-based PWM scheme has been
performance compared to the SPWM schemes [3–5]. The presented [15], where sinusoidal references are added with a
maximum peak of the fundamental component in the output proper offset voltage before being compared with carriers, to
voltage obtained with space vector modulation is 15% greater achieve the performance of a SVPWM. The offset voltage
than with the sine-triangle modulation scheme [2, 3]. But the computation is based on a modulus function depending on
conventional SVPWM scheme requires sector identification the DC link voltage, number of levels and the sinusoidal
and look-up tables to determine the timings for various reference signal amplitudes. A SVPWM scheme is presented
switching vectors of the inverter, in all the sectors [3, 4]. This [18], where the switching time for the inverter legs is directly
makes the implementation of the SVPWM scheme quite determined from sampled sinusoidal reference signal
complicated. It has been shown that, for two-level inverters, amplitudes for five-level inverter where two three-level
a SVPWM like performance can be obtained with a SPWM inverters feed the dual-fed induction motor. The objective of
scheme by adding a common mode voltage of suitable this paper is to present an implementation scheme for PWM
magnitude, to the sinusoidal reference signals [4, 5]. A signal generation for five-level inverter system for dual-fed
simplified method, to determine the correct offset times for induction motor, similar to the SVPWM scheme. In the
centering the time durations of the middle space vectors, in a proposed scheme, the dual-fed induction motor is fed with
sampling time interval, is presented [8], for the two-level four-level inverter from one end and two-level inverter from
47
© 2011 ACEEE
DOI: 01.IJEPE.02.02. 17


the other end, four-level inversion is obtained by connecting the five levels generated for phase-A are shown in TABLE I.
three conventional 2- level inverters with equal DC link TABLE I
voltage in cascade. The PWM switching times for the inverter THE FIVE LEVELS REALIZED IN THE PHASE-A WINDING
legs are directly derived from the sampled amplitudes of the
sinusoidal reference signals. A simple way of adding an offset
voltage to the sinusoidal reference signals, to generate the
SVPWM pattern, from only the sampled amplitudes of
sinusoidal reference signals, is explained. The proposed
SVPWM signal generation does not involve checks for region
identification, as in the SVPWM scheme presented in [14].
Also, the algorithm does not require either sector
identification or look-up tables for switching vector III. VOLTAGE SPACE VECTORS OF PROPOSED
determination as are required in the conventional multilevel INVERTER
SVPWM schemes [10,11]. Thus the scheme is At any instant, the combined effect of 1200 phase shifted
computationally efficient when compared to conventional three voltages in the three windings of the induction motor
multilevel SVPWM schemes, making it superior for real-time could be represented by an equivalent space vector. This
implementation. space vector Es, for the proposed scheme is given by
Es=EA3A4+EB3B4.ej(2  /3) + EC3C4 . ej(4  /3). (1)
II. FIVE-LEVEL INVERTER SCHEME FOR THE DUAL-FED
By substituting expressions for the equivalent phase
INDUCTION MOTOR
voltages in (1)
The power circuit of the proposed drive is shown in Fig.1. Es = (EA3n – EA4n’) + (EB3n – EB4n’). ej(2/3) + (EC3n – EC4n’). ej(4/3)
Four-level inverter from one end and two-level inverter from (2)
the other end feed the dual-fed induction motor. The four- This equivalent space vector Es can be determined by
level inverter is composed of three conventional two-level resolving the three phase voltages along mutually
inverters INV-1, INV-2 and INV-3 in cascade. The DC link perpendicular axes, d-q axes of which d-axis is along the A-
voltage of INV-1, INV-2, INV-3 and the two-level inverter INV- phase (Fig.2). Then the space vector is given by
4 is (1/4)Edc, where Edc is the DC link voltage of an equivalent
Es = Es (d) +j Es(q) (3)
conventional single two-level inverter drive. The leg voltage
EA3n of phase-A attains a voltage of (1/4)Edc if (i)The top Where Es(d) is the sum of all voltage components of EA3A4,
switch S31 of INV-3 is turned on (Fig.1) and (ii) The bottom EB3B4 and EC3C4 along the d-axis and Es(q) is the sum of the
switch S24 of INV-2 is turned on. The leg voltage EA3n of voltage components of EA3A4, EB3B4 and EC3C4 along the q-axis.
phase-A attains a voltage of (2/4)Edc if (i) the top switch S31 The voltage components Es(d) and Es(q) can be thus
of INV-3 is turned on (ii) The top switch S21 of INV-2 is turned expressed by the following transformation,
on and (iii) The bottom switch S14 of INV-1 is turned on. The
ES(d) = EA3A4(d) +EB3B4 (d) +EC3C4 (d) (4)
leg voltage EA3n of phase-A attains a voltage of (3/4)Edc if
(i)The top switch S31 of INV-3 is turned on (ii) The top switch
ES(q) = EB3B4(q) + EC3C4(q) (5)
S21 of INV-2 is turned on and (iii)The top switch S11 of INV-1
is turned on. The leg voltage EA3n of phase-A attains a voltage
of zero volts if the bottom switch S34 of the INV-3 is turned  1 1 
E
on. Thus the leg voltage EA3n attains four voltages of 0, (1/  Es (d ) 1  2  2   A3 A 4 
4)Edc, (2/4)Edc and (3/4)Edc, which is basic characteristic of  Es (q)     B3 B 4 
E
a 4-level inverter. Similarly the leg voltages EB3n and EC3n of    3 3   (6)
 0   E C 3C 4 
 
phase-B and phase-C attain the four voltages of 0, (1/4)Edc,  2 2 
(2/4)Edc and (3/4)Edc. The leg voltage EA4n’ of phase-A attains
a voltage of (1/4)Edc if the top switch S41 of INV-4 is turned By substituting expressions for the equivalent phase voltages
on . The leg voltage EA4n’ of phase-A attains a voltage of zero in (6),
volts if the bottom switch S44 of the INV-4 is turned on. Thus
the leg voltage EA4n’ attains two voltages of 0 and (1/4)Edc,  1 1 
E  E A4 n' 
which is basic characteristic of a 2-level inverter. Similarly the  Es (d ) 1  2  2   A3n
E  E 
leg voltages EB4n’ and EC4n’ of phase-B and phase-C attain the  Es (q)     B 3n
3 
B 4 n' 
   3 (7)
two voltages of 0 and (1/4)Edc. Thus, one end of dual-fed 0   EC 3n  EC 4 n' 
  
induction motor may be connected to a DC link voltage of  2 2 
either zero or (1/4)Edc or (2/4)Edc or (3/4)Edc and other end
may be connected to a DC link voltage of either zero or (1/
4)Edc. When both the inverters, four-level inverter and two-
level inverter drive the induction motor from both ends, five
different levels are attained by each phase of the induction
motor. If we assume that the points n and n’ are connected,
48
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


Fig.1 Schematic circuit diagram of the proposed 5- level inverter drive scheme

The inverters can generate different levels of voltage vectors
in the three phases of induction motor depending upon the
condition of the switchings of inverter and for each of the
different combinations of leg voltages, EA3n, EB3n and EC3n for
the four-level inverter and EA4n’, EB4n’ and EC4n’ for the two-
level inverter. The different equivalent voltage space vectors
can be determined using (3) and (7). The possible
combinations of space vectors will occupy different locations
as shown in Fig. 3. There are in total 61 locations forming 96
sectors in the space vector point of view. The resultant
hexagon (Fig.3) can be divided into four layers: layer-
1(innermost layer); layer-2(next outer layer); layer-3(layer
outside layer 2) and layer-4(outermost layer).

Fig 3. The voltage space vector locations and layers for the
proposed drive

IV. EFFECT OF COMMON-MODE VOLTAGE IN SPACE
VECTOR LOCATIONS
In the above analysis to generate the different levels and
the space vector locations, the points n and n’ are assumed
to be connected. When the points n and n’ are not connected
(as in the proposed topology Fig.1), the actual motor phase
voltages are
Fig.2 Determination of equivalent space vector from phase EA3A4 = EA3n – EA4n’ – En’n (8)
voltages
EB3B4 = EB3n – EB4n’ – En’n (9)

EC3C4 = EC3n – EC4n’ – En’n (10)
En’n is the common-mode voltage and is given by

49
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


sinusoidal reference signals cross the triangular carrier signals
1 at different instants in a sampling time interval Ts (Fig.4).
En’n = (EA3n + EB3n + EC3n) –
3 Each time a sinusoidal reference signal crosses the triangular
carrier signal, it causes a change in the inverter switching
1 state. The changes in phase voltage and their time intervals
(E + E + E ) (11)
3 A4n’ B4n’ C4n’ are shown in Fig.5 in a sampling time interval Ts. The sampling
Substituting these expressions in (1) time interval Ts, can be split into four time intervals t01, t1,t2
and t02. The time intervals t01and t02 are the time durations for
ES = (EA3n – EA4n’ – En’n) + (EB3n – EB4n’ – En’n). ej(2  /3)
+ (EC3n – the start and end inverter space vectors respectively, in a
EC4n’ – En’n). ej(4/3) = (EA3n – EA4n’) + sampling time interval Ts. The time intervals t1 and t2 are the
(EB3n – EB4n’). ej(2/3) + (EC3n – EC4n’). ej(4/3) - (En’n + time durations for the middle inverter space vectors (active
En’n . ej(2/3) + En’n . ej(4/3)) space vectors), in a sampling time interval Ts. It should be
In this equation observed from Fig.5 that the middle space vectors are not
(En’n + En’n . ej(2/3) + En’n . ej(4/3)) = En’n - ½ En’n -½ En’n = 0 centered in a sampling time interval Ts. Because of the level-
and the equation then reduces to shifted four triangular carrier signals (Fig.4), the first crossing
(termed as first_cross) of the sinusoidal reference signal
Es = (EA3n – EA4n’) + (EB3n – EB4n’). ej(2/3) + cannot always be the minimum magnitude of the three
(EC3n – EC4n’). ej(4/3) sampled sinusoidal reference signals, in a sampling time
This expression of Es is the same as (2), where the points n interval. Similarly, the last crossing (termed as third_cross)
and n’ are assumed to be connected. The above analysis of the sinusoidal reference signal cannot always be the
depicts that the common-mode voltage present between the maximum magnitude of the three sampled sinusoidal reference
points n and n’ does not effect the space vector locations. signals, in a sampling time interval. Thus the offset voltage,
This common-mode voltage will effect only in the diversity Eoffset1 is not sufficient to center the middle inverter space
of space vectors in different locations. vectors, in a multilevel PWM system during a sampling time
interval Ts (Fig.5). Hence an additional offset (offset2) has to
V. PROPOSED SVPWM IN LINEAR MODULATION be added to the sinusoidal reference signals of Fig.4, so that
RANGE the middle inverter space vectors can be centered in a
For two-level inverters, in the SPWM scheme, each sampling time interval, same as a two-level SVPWM system
sinusoidal reference signal is compared with the triangular [3].In this paper, a simple procedure to find out the offset
carrier signal and the individual phase voltages are generated voltage (to be added to the sinusoidal reference signals for
[1]. To attain the maximum possible peak amplitude of the PWM generation) is presented, based only on the sampled
fundamental phase voltage, a common offset voltage, Eoffset1 amplitudes of the sinusoidal reference signals. In the proposed
is added to the sinusoidal reference signals [5, 12], where the scheme, the sinusoidal reference signal, from the three
magnitude of Eoffset1 is given by sampled sinusoidal reference signals, which crosses the
triangular carrier signal first (first_cross) and the sinusoidal
Eoffset1= -- (Emax + Emin) / 2 (12) reference signal which crosses the triangular carrier signal
Where Emax and Emin are the maximum and minimum last (third_cross) are found. Once the first_cross signal and
magnitudes of the three sampled sinusoidal reference signals third_cross signal are known, the theory of offset calculation
respectively, in a sampling time interval. The addition of this of (12), for the 2-level inverter, can easily be adapted for the
common offset voltage, Eoffset1, results in the active space 5-level SVPWM generation scheme.
vectors being centered in a sampling time interval, making
the SPWM scheme equivalent to the SVPWM scheme [3]. In
a sampling time interval, the sinusoidal reference signal which
has lowest magnitude crosses the triangular carrier signal
first, and causes the first transition in the inverter switching
state. While the sinusoidal reference signal, which has the
maximum magnitude, crosses the triangular carrier signal last
and causes the last switching transition in the inverter
switching states in a two-level SVPWM scheme [5, 13]. Thus
the switching times of the active space vectors can be
determined from the sampled sinusoidal reference signal
amplitudes in a two-level inverter system [8]. The SPWM
scheme, for five-level inverter, sinusoidal reference signals
are compared with symmetrical level shifted four triangular
carrier signals for PWM generation [9]. After addition of
offset voltage Eoffset1 to the sinusoidal reference signals,
the modified sinusoidal reference signals are shown in Fig.4 Fig. 4 Modified sinusoidal reference signals and triangular carrier
along with four triangular carrier signals T1 to T4. The signals for a five-level PWM scheme
50
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


Where T*as, T*bs and T*cs are the time equivalents of the
voltage magnitudes. The proportionality between the time
equivalents and corresponding voltage magnitudes is defined
as follows [8]:
(Edc “ 4) “Ts = E*AN “ T*as
(Edc “ 4) “Ts = E*BN “ T*bs
(Edc “ 4) “Ts = E*CN “ T*cs
(Edc “ 4) “Ts = Eoffset1 “ Toffset1 (15)
The time interval, at which the sinusoidal reference signals
cross the triangular carrier signals for the first time, is termed
as Tfirst_cross. Similarly, the time intervals, at which the
sinusoidal reference signals cross the triangular carrier signals
for the second and third time, are termed as, Tsecond_cross
Fig. 5 Inverter switching vectors and their switching time durations and Tthird_cross respectively, in a sampling time interval Ts.
during sampling time interval Ts
Tfirst_cross = min (Ta-cross, Tb-cross, Tc-cross)
VI. DETERMINATION OF THE OFFSET VOLTAGE FOR A Tsecond_cross = mid (Ta-cross, Tb-cross, Tc-cross)
FIVE-LEVELINVERTER Tthird_cross = max (Ta-cross, Tb-cross, Tc-cross) (16)

Fig.4 shows modified sinusoidal reference signals and The time intervals, Tfirst_cross, Tsecond_cross and
four triangular Carrier signals used for PWM generation for Tthird_cross, directly decide the switching times for the
five-level inverter. The modified sinusoidal reference signals different inverter voltage vectors, forming a triangular sector,
are given by during one sampling time interval Ts. The time intervals for
E*AN=EAN + Eoffset1 the start and end space vectors, are t01= Tfirst_cross,
E*BN=EBN + Eoffset1 t02= (Ts “ Tthird_cross), respectively (Fig.5). The middle
E*CN=ECN + Eoffset1 (13) space vectors are centered by adding a time offset, Toffset2
where EAN, EBN and ECN are the sampled amplitudes of to Tfirst_cross, Tsecond_cross and Tthird_cross. The time
sinusoidal reference signals during the current sampling time offset, Toffset2 is determined as follows. The time interval
interval and Eoffset1 is calculated from (12).The time interval, for the middle inverter space vectors, Tmiddle, is given
at which the A-phase sinusoidal reference signal, E*AN crosses by:
the triangular carrier signal, is termed as Ta-cross (Fig.6). Tmiddle= Tthird_cross --Tfirst_cross (7)
Similarly, the time intervals, when the B-phase and C-phase
sinusoidal reference signals, E*BN and E*CN cross the triangular The time interval of the start and end space vector is
carrier signals, are termed as Tb-cross and Tc-cross T0 =Ts --Tmiddle (18)
respectively. Fig.6 shows a sampling time interval when the
A-phase sinusoidal reference signal is in the triangular carrier Thus the time interval of the start space vector is given by
region T3 while the B-phase sinusoidal reference signal and T0 /2= Tfirst_cross + Toffset2
C-phase sinusoidal reference signal are in carrier region T4 Therefore
and T2 respectively. As shown in Fig.6, the time interval, Ta-
cross, at which the A-phase sinusoidal reference signal Toffset2 = T0 / 2 --Tfirst_cross (19)
crosses the triangular carrier signal is directly proportional
to the phase voltage amplitude, (E*AN “ Edc “4). The time
interval, Tb-cross, at which the B-phase sinusoidal reference
signal crosses the triangular carrier signal, is proportional to
(E*BN + Edc “2) and the time interval, Tc-cross, at which the C-
phase sinusoidal reference signal crosses the triangular car-
rier signal, is proportional to (E*CN + Edc “4). Therefore

Fig.6 Determination of the Ta-cross, Tb-cross and Tc-cross during
sampling interval Ts
In this way, we can obtain offset voltages to be added for
remaining samples during the time period of sinusoidal refer-
ence signal. For 5-level inverter maximum modulation index
in the linear modulation range is 0.866 (the modulation index,
M, is defined as the ratio of magnitude of the equivalent refer-
ence voltage space vector, generated by the three sinusoidal
51
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


reference signals, to the DC link voltage). The proposed normalized harmonic spectrum of motor phase voltage are
scheme can be adapted for modulation indices lesser than shown in Fig.10d and Fig.10e respectively. The ratio of
0.866 .The addition of the time offset, Toffset2 to Ta- triangular carrier signal frequency to reference sinusoidal
cross, Tb-cross and Tc-cross gives the inverter leg switch- signal frequency is 48 for all ranges of operation. It can be
ing times Tga, Tgb and Tgc for phases A, B and C, respec- observed that the motor phase voltage and motor phase
tively. current during 5-level operation are very smooth and close
to the sinusoid with lower harmonics.
Tga =Ta-cross + Toffset2
Tgb = Tb-cross + Toffset2
Tgc = Tc-cross + Toffset2 (20)

VII. SIMULATION RESULTS AND DISCUSSION
The proposed SVPWM scheme is simulated using
MATLAB environment with open loop E/f control for
different modulation indices. The DC link voltage applied is
(1/4)Edc for the INV-1, INV-2, INV-3 and INV-4, where Edc is
the DC link voltage of an equivalent conventional single two-
level inverter drive. The speed reference is translated to the
Fig.7a. Motor phase voltage when M=0.15 (layer-1, 2-level
frequency and voltage commands maintaining E/f. The opera tion)
modified three reference sinusoidal signals which are added
by the total offset voltage to make SPWM scheme equivalent
to the SVPWM scheme, are simultaneously compared with
the triangular carrier set. A DC link voltage (Edc) of 400 volts
is assumed for simulation studies. Fig.7a shows the motor
phase voltage (E A3A4) in the lowest speed range which
corresponds to layer-1 operation (two-level mode) when
modulation index is 0.15. Fig.7b shows the total offset voltage
to be added to sinusoidal reference signals to make SPWM
equivalent to the SVPWM and Fig.7c shows the A-phase
sinusoidal reference signal after offset voltage is added.
During this range of operation, motor phase current and Fig.7b. The offset voltage to be added to sinusoidal reference
normalized harmonic spectrum of motor phase voltage are signals when M=0.15 (layer-1, 2-level operation)
shown in Fig.7d and Fig.7e respectively. Fig.8a shows the
motor phase voltage (EA3A4) in the next speed range which
corresponds to layer-2 operation (three-level mode) when
During this range of operation, motor phase current and
normalized harmonic spectrum of motor phase voltage are
Fig.7c. The A-phase sinusoidal reference signal after offset voltage
motor phase voltage (EA3A4) in the next speed range which is added when M=0.15 (layer-1, 2-level operation)
corresponds to layer-3 operation (four-level mode) when
normalized harmonic spectrum of motor phase voltage are
motor phase voltage (EA3A4) in the next speed range which
corresponds to layer-4 operation (five-level mode) when Fig.7d. Motor phase current when M=0.15 (layer-1, 2-level
modulation index is 0.85. Fig.10b shows the total offset opera tion)
voltage to be added to sinusoidal reference signals to make
SPWM equivalent to the SVPWM and Fig.10c shows the A-
phase sinusoidal reference signal after offset voltage is added.
52
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


Fig.8e. Normalized harmonic spectrum of the motor phase voltage
Fig.7e. Normalized harmonic spectrum of the motor phase voltage when M=0.3 (layer-2, 3-level operation)
when M=0.15 (layer-1, 2-level operation)

Fig.9a. Motor phase voltage when M=0.6 (layer-3, 4-level
Fig.8a. Motor phase voltage when M=0.3 (layer-2, 3-level opera tion)
opera tion)

Fig.9b. The offset voltage to be added to sinusoidal reference
Fig.8b. The offset voltage to be added to sinusoidal reference signals when M=0.6 (layer-3, 4-level operation)
signals when M=0.3 (layer-2, 3-level operation)

is added when M=0.3 (layer-2, 3-level operation)
is added when M=0.6 (layer-3, 4-level operation)

Fig.8d. Motor phase current when M=0.3 (layer-2, 3-level
opera tion) Fig.9d. Motor phase current when M=0.6 (layer-3, 4-level
opera tion)

53
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


Fig.10e. Normalized harmonic spectrum of the motor phase
Fig.9e. Normalized harmonic spectrum of the motor phase voltage voltage when M=0.85 (layer-4, 5-level operation)
when M=0.6 (layer-3, 4-level operation)
VIII. CONCLUSIONS
A modulation scheme of SVPWM for dual-fed induction
motor drive, where the induction motor is fed by four-level
inverter from one end and two-level inverter from the other
end is presented. The four-level inverter used is composed
of three conventional two-level inverters with equal DC link
voltage in cascade. The centering of the middle inverter space
vectors of the SVPWM is accomplished by the addition of
an offset voltage signal to the sinusoidal reference signals,
derived from the sampled amplitudes of the sinusoidal
Fig.10a. Motor phase voltage when M=0.85 (layer-4, 5-level reference signals. The SVPWM technique, presented in this
opera tion) paper does not require any sector identification, as is required
in conventional SVPWM schemes. The proposed scheme
eliminates the use of look-up table approach to switch the
appropriate space vector combination as in conventional
SVPWM schemes. This reduces the computation time
required to determine the switching times for inverter legs,
making the algorithm suitable for real-time implementation.

REFERENCES
[1] Holtz, J.: ‘Pulsewidth modulation–A survey’, IEEE Trans. Ind.
Fig.10b. The offset voltage to be added to sinusoidal reference Electron., 1992, 30, (5), pp. 410–420
signals when M=0.85 (layer-4, 5-level operation) [2] Zhou, K., and Wang, D.: ‘Relationship between space-vector
modulation and three-phase carrier-based PWM: A comprehensive
analysis’, IEEE Trans. Ind. Electron., 2002, 49, (1), pp. 186–196
[3] Van der Broeck, Skudelny, H.C., and Stanke, G.V.: ‘Analysis
and realisation of a pulsewidth modulator based on voltage space
vectors’, IEEE Trans. Ind. Appl., 1988, 24, (1), pp. 142–150
[4] Boys, J.T., and Handley, P.G.: ‘Harmonic analysis of space
vector modulated PWM waveforms’, IEE Proc. Electr. Power
Appl., 1990, 137, (4), pp. 197–204
[5] Holmes, D.G.: ‘The general relationship between regular sampled
pulse width modulation and space vector modulation for hard
switched converters’. Conf. Rec. IEEE Industry Applications
Fig.10c. The A-phase sinusoidal reference signal after offset
Society (IAS) Annual Meeting, 1992, pp. 1002–1009
voltage is added when M=0.85 (layer-4, 5-level operation)
[6] Holtz, J., Lotzkat, W., and Khambadkone, A.: ‘On continuous
control of PWMinverters in over-modulation range including six-
step mode’, IEEE Trans. Power Electron., 1993, 8, (4), pp. 546–
553
[7] Lee, D., and Lee, G.: ‘A novel overmodulation technique for
space vector PWM inverters’, IEEE Trans. Power Electron., 1998,
13, (6), pp. 1144–1151
[8] Kim, J., and Sul, S.: ‘A novel voltage modulation technique of
the Space Vector PWM’. Proc. Int. Power Electronics Conf.,
Yokohama, Japan, 1995, pp. 742–747
Fig.10d. Motor phase current when M=0.85 (layer-4, 5-level
[9] Carrara, G.,Gardella, S.G., Archesoni,M., Salutari, R., and
opera tion) Sciutto, G.: ‘A new multi-level PWM method: A theoretical
analysis’, IEEE Trans. Power Electron., 1992, 7, (3), pp. 497–505
54
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17


[10] Shivakumar, E.G., Gopakumar, K., Sinha, S.K., Andre, Pittet, [15] McGrath, B.P., Holmes, D.G., and Lipo, T.A.: ‘Optimized
and Ranganathan, V.T.: ‘Space vector PWM control of dual inverter space vector switching sequences for multilevel inverters’. Proc.
fed open-end winding induction motor drive’. Proc. Applied IEEE Applied Power Electronics Conf. (APEC), 2001, pp. 1123–
PowerElectronics Conf. (APEC), 2001, pp. 339–405 1129
[11] Suh, J., Choi, C., and Hyun, D.: ‘A new simplified space [16] Krah, J., and Holtz, J.: ‘High performance current regulation
vector PWM method for three-level inverter’. Proc. IEEE Applied and efficient PWM Implementation for low inductance servo
Power Electronics Conf. (APEC), 1999, pp. 515–520 motors’, IEEE Trans. Ind. Appl., 1999, 36, (5), pp. 1039–1049
[12] Baiju, M.R., Mohapatra, K.K., Somasekhar, V.T., Gopakumar, [17] Somasekhar, V.T., and Gopakumar, K.: ‘Three-level inverter
K., and Umanand, L.: ‘A five-level inverter voltage space phasor configuration cascading two 2-level inverters’, IEE Proc. Electr.
generation for an open-end winding induction motor drive’, IEE Power Appl., 2003, 150, (3), pp. 245–254
Proc. Electr. Power Appl., 2003, 150, (5), pp. 531–538 [18] R.S.Kanchan, M.R.Baiju, K.K.Mohapatra, P.P.Ouseph and
[13] Wang, FEI: ‘Sine-triangle versus space vector modulation for K.Gopakumar, ‘Space vector PWM signal generation for multilevel
threelevel PWM voltage source inverters’. Proc. IEEE-IAS Annual inverters using only the sampled amplitudes of reference phase
Meeting, Rome, 2000, pp. 2482–2488 voltages’ IEE Proc.-Electr. Power Appl., Vol. 152, No. 2, March
[14] Baiju, M.R., Gopakumar, K., Somasekhar, V.T., Mohapatra, 2005, pp.297-309.
K.K., and Umanand, L.: ‘A space vector based PWMmethod using [19] K.Chandra Sekhar and G.Tulasi Ram Das ‘Five-level SPWM
only the instantaneous amplitudes of reference phase voltages for Inverter for an Induction Motor with Open-end windings’ PECon
three-level inverters’, EPE J., 2003, 13, (2), pp. 35–45 2006, Putrajaya, Malaysia, November 28-29, 2006,pp.342-347.

55
© 2011 ACEEE
DOI: 01.IJEPE.02.02.17

A Refined Space Vector PWM Signal Generation for Multilevel Inverters

More Related Content

What's hot (20)

Similar to A Refined Space Vector PWM Signal Generation for Multilevel Inverters (20)

More from IDES Editor (20)

Recently uploaded (20)

A Refined Space Vector PWM Signal Generation for Multilevel Inverters