Simulation of Five-level Five-phase SVPWM

Voltage Source Inverter

M.M.Irfan, P.H.K.Prasad, .and P.Venugopal Rao,

Abstract-- Multilevel voltage-fed inverters with space vector pulse

width modulation strategy are gained importance in high power II. REVIEW OF PREVIOUS RESEARCH
high performance industrial drive applications. Multi-phase Three-phase voltage source inverters are widely used in
machines and drives is a topic of growing relevance in recent
variable speed ac motor drives applications since they provide
years, and it presents many challenging issues that still need
further research. Multilevel multiphase technology combines the variable voltage and variable frequency output through pulse
benefits of both technologies. This paper proposes a new width modulation control. Continuous improvement in terms
simplified space vector PWM method for a five-level five phase of cost and high switching frequency of power semiconductor
voltage source inverter. The five- level inverter has a large devices and development of machine control algorithm leads
number of switching states compared to a two-level inverter. In to growing interest in more precise PWM techniques.
the proposed scheme, five-level space vector PWM inverter is
Volume of work has been carried out in this direction
easily implemented using algorithm based on a displacement plus
a two-level multiphase SVPWM modulator. Therefore, the and a review of popular techniques are presented in by
proposed method can also be applied to multilevel inverters. In Holtz (1992) and Holtz (1994). The most widely used
this paper, a five-level five phase inverter using space vector PWM method is the carrier-based sine-triangle PWM
modulation strategy has been modeled and simulated. Simulation method due to simple implementation in both analog
results are presented for operation conditions using R-L load. and digital realization. However, the dc bus utilization
in this method is low 0(.5 Vdc ) . This has led to the
Index Terms—Decomposition, Multilevel Inverter, Multi
Phase Machine, Svpwm
investigation into other techniques with an objective of
improvement in the dc bus utilization. It is found by
I. INTRODUCTION Houdsworth and Grant (1984) that injection of zero
sequence (third harmonic) extends the range of operation
T his Multilevel converter technology is based on the
synthesis of a Voltage waveform from several dc voltage
levels. As the number of levels increases, the synthesized
of modulator by 15.5%. The major problem associated with
high power drive applications is high switching losses in
inverters. To reduce switching losses a PWM technique
output voltage gets more Steps and produces a waveform
termed as discontinuous PWM (DPWM) was developed by
which approaches the reference More accurately. The major
Depenbrock (1977) and Kolar et al. (1991). The proposed
advantages of using multi level Inverters are:
discontinuous PWM techniques were based on triangle-
1) high voltage capability with voltage limited devices;
intersection-implementation, in which non sinusoidal
2) low harmonic distortion;
modulating signal is compared with triangular carrier
3) reduced switching losses;
wave. A generalized discontinuous PWM algorithm was
4) increased efficiency;
presented by Hava et al.(1998) which encompasses
5) good electromagnetic compatibility.
the techniques presented by Depenbrock (1977) and
The interest in multi-phase motor drives has increased in
Kolar et al. (1991). However, the better visualization of the
recent years due to several advantages when compared to
DPWM methods is obtained by using space vector theory. A
three-phase drives [1-2]. Some of these advantages are known
generalized DPWM based on space vector theory is available
from the early days of the multi-phase drives [3], although
in Grahame and Lipo (2000), Kazmierkowski et al. (2002),
they have recently considered with high-level analysis [4].
Dong (2005) and Hava (1998). The discontinuous PWM
These advantages are inherent to the own structure of the
strategies is based on the principle of eliminating one of the
machine, and include less torque ripple, less acoustic noise
zero voltage vectors, causing the active voltagespace vectors
and losses, reduced current per phase, greater efficiency and
to join together in two successive half switching interval. The
increased reliability due to the additional number of phases [1-
discontinuous PWM techniques have the advantage of
4]. These advantages make multi-phase drives suitable for
eliminating one switching transition in each half switching
high power/current applications such as EV/HEVs traction or
period, consequently reducing the number of switching by one
electric ship propulsion.
third. alternatively, the switching frequency can be increased
M.M.Irfan, isassociated with ,ADAM’S Eng Colleg ,Dr.P.H.K.Prasad,Prof. &
HOD, EEE, ADAM’S Engg College ; Prof.P.Venugopal Rao, HOD EEE, S.R by 3/2 for the same inverter losses. Another PWM technique
Engg College termed as Space vector PWM based on space vector theory
was proposed by de Broeck et al. (1988) and Ogasawara et al.
(1989) which offers superior performance compared to the

978-1-4244-8542-0/10/$26.00 ©2010 IEEE

 2

carrier-based sine-triangle PWM technique in terms of higher Where the sum of the intervals Tj must be equal tothe
dc bus utilization and better harmonic performance. Further modulation period T. The reference vector summarizes the
this technique offers easier digital realization.A voltage reference for each Phase of the system, whereas each
comprehensive relationship between the carrier-based and switching vector summarizes. The switching state of each
SVPWM is illustrated by Holmes (1996) and it was phase of the converter
realized that the placement of the active and zero space
vectors in each half switching period is the only difference
between the carrier-based scheme and SVPWM. With the
advancement in the power semiconductor devices the
switching frequency has increased many folds. For high Therefore ,the reference vector and the switching vectors
switching frequency applications, Artificial Neural Network belong to the multidimensional space Rp ,where P is the
(ANN) based space vector PWM technique is developed in number of phases of the converter. In most common
Iqbal et al. (2006), Pinto et al. (2000), Kerkmanet al. multilevel topologies such as flying capacitor, diode-clamped,
(1991), Haykin (2004), and Muthuramalingam and Himavathi cascaded full-bridge or hybrid converters, the output level of
(2007).The implementation of ANN based SVPWM is every phase Vs is an integer multiple of a fixed voltage step
fast and offers higher bandwidth of control loop and Vdc [5],[7]
reduced harmonics. The back propagation feed forward
network is implemented for gate signal generation required for
SVPWM.
And Recent multilevel multiphase papers are International therefore, vectors and switching times can be normalized by
Journal on Electrical Engineering and Informatics - Volume using the voltage step and the switching period ,respectively to
2, Number 2, 2010 ,137 “Performance Improvement of non dimensionalize (1)and(2)
Multiphase Multilevel Inverter Using Hybrid Carrier Based
Space Vector Modulation “C. Govindaraju1 and K.
Baskaran2; “Multi-Phase Space Vector Pulse Width
Modulation: Applications and strategies”. Mario J. Duran1, F.
Barrero2 and S. Toral; “A Simple Approach to Space Vector
PWM Signal Generation for a Five-Phase Voltage Source
Inverter” Mohd. Arif Khan, Atif Iqbal ; “Multilevel
Multiphase Space Vector PWM Algorithm Applied to
Three-Phase Converters”´Oscar Lopez,, Jacobo ´ Alvarez, It is important to remark that new normalized switching
Jes. with regard to multilevel multiphase SVPWM, an vectors vsj now belong to the multidimensional space of
algorithm for a neutral clamped five-phase inverter was integer numbers ZP.If the above expressions are substituted
proposed in [6].However, it does not address the extension of in(1)and(2) ,the modulation law can be rewritten interms of
the method for a higher Number of levels or phases or its the new Normalized variables as
application to other multilevel topologies. In this paper, a
generic algorithm to perform the SVPWM for multiphase
inverters is presented. This algorithm, which is valid for the
typical multilevel topologies.

III ALGORITHM FORMULATION

Since the switching states of any power converter topology
Stay at discrete states, the SVPWM issued to approximate If the reference and the switching normalized vectors are
Reference voltage vector Vr by means of a sequence of space Expressed as follows:
vectors S1 = {Vs1,Vs2,...,Vsl} during each modulation cycle.
To achieve a proper synthesis of the reference vector ,each
switching vector Vsj must be applied during an interval Tj in
accordance with the following modulation law:
 3

The above system of linear equations constitutes the

modulation law, which must be solved by the multilevel
multiphase SVPWM algorithm. The problem solving includes
three main steps:
1) searching a set of integer coefficients for the matrix
that Permits solving the linear system;
2) solving the system of linear equations to calculate the This new system of line are quations presents the same
Switching times; form As the general modulation law (13). However, in this
3) extracting the switching vector sequence from the case, the Components of vector vf are bounded in the interval
coefficient matrix. [0, 1). Therefore, only the subset of displaced vectors with
The multilevel multiphase SVPWM problem can be compoNents zero or one is enough to carry out the reference
simplified if it is decomposed into the sum of a displacement approximation. Consequently, this new equation represents a
plus a two-level SVPWM problem with the same number of two-level modulator where the reference vector is vf and the
phases. array of switching vectors are the displaced set of switching
A. Algorithm Decomposition vectors vdj .Switching times are the same in the multilevel and
The reference vector can be decomposed into the sum of its the two-level modulators. Fig.1shows a 2-D example of the
Integer and fractional parts decomposition where vector vi coincides with a switching
vector and the subset {[0, 0], [1, 0], [0, 1], [1, 1]} of displaced
vectors is enough to synthesize the fractional part of the
reference vf .
Components of the new vector vi are integer numbers, and
Therefore it belongs to the same space ZP of the switching
Vectors and it could be directly synthesized with one of them.
The fractional part vr still belongs to the space RP and it
cannot be directly synthesized by means of a single switching
vector. It has to be approximated with a sequence of switching
vectors. Besides , a new set of switching vectors is obtained
by displacing all switching vectors the distance given by vi

In summary,(19)demonstrates that a multilevel multiphase

If those vectors are expressed as Modulator can be realized from a displacement plus a two-
level Modulator with the same number of phases.Fig.2shows
a Block diagram of the proposed technique.
B. Two-Level Multiphase SVPWM Algorithm
Once the multilevel problem has been decomposed, the
two-Level modulation law (21) has to be solved. To obtain an
And if(15)is substituted in(13),the new expression for the exactly determined system of linear equations, the coefficient
modulation law is obtained matrix of that modulation law must be a square matrix. Hence,
the length of the switching vector sequence Sl must be
l = P +1 (22)
and the particular linear system, which has to be solved ,is

The objective of the two-level modulation algorithm is to

find a Switching vector sequence ,that is ,the coefficient
matrix of the System (23)should be filled with zeros and ones,
thus allowing A subsequent system solution.Moreover,the
And if (19) and (20) are compared, the following
coefficient selection must be carried out taking into account
relationship between the fractional part of the reference and
that the switching Times must be always positive after the
the displaced Switching vectors is obtained:
system solution.
 4

There are many different possibilities to fill the coefficient and, consequently, the number of ones and zeros in each
matrix. Nevertheless, the whole power system performance column does not change. Hence, the switching number is
depends on the method employed for calculating the minimized because consecutive vectors of the sequence are
coefficient matrix. In this way, switching losses are minimized still adjacent after the transformation.
if coefficients are selected in such a way that consecutive Due to the fact that the solution t is the same for both
switching vectors of the switching sequence are adjacent. In linear systems,(24)and(27),it can be calculated by using either
other words, only one coefficient is different in two of them. The second option seems the best choice because, in
consecutive matrix columns. One possible method for this case, the solution is trivial as shown below
calculating such a matrix is detailed below.
Equation (23)can be written in a shorter form as

Finding a permutation matrix P that puts the elements of the All equations calculated by means of the above expression
Reference vector vf in descending order will Always be positive numbers because the coordinates of
the vector ˆ vf obey(26). In summary, matrix D permits
solving the two-level modulation law getting positive
switching times and minimizing Switching number. The two-
level switching sequence can be Directly extracted from the
Where columns of that matrix.

And multiplying both sides of (24)by this permutation matrix IV SIMULATION MODEL
P, we obtain the following equation:

where

One coefficient matrix ˆ D with adjacent consecutive

columns That makes this new system of linear equations
exactly determined is the following upper triangular matrix:

Figure. Simulink model for 5level 5 phase svpwm inverter

Circuit description:
This model illustrates the operation of the blocks
As it will be shown below, the switching times obtained multilevel multiphase SVPWM and Two-level multiphase
with This coefficient matrix are always positive. A SVPWM. It includes a five-level five-phase inverter feeding a
permutation matrix is an orthogonal matrix so P is invertable passive load.
and Therefore, the coefficient matrix D of the two-level 1. The 'Sine Wave f' and 'Fifth harmonic' blocks generate
modulation Law can be obtained by solving (28) as an unbalanced five-phase reference voltage with a fifth
harmonic.
2. The 'Normalize' block normalize the reference voltage
3. The 'Multilevel multiphase SVPWM' performs the
multilevel multiphase space-vector PWM (SVPWM)
algorithm. This block makes use of the block 'Two-level
The permutation matrix P applies a set of elementary row multiphase SVPWM, that can be used alone with two-level
Switching transformations to the column vector vf . In the multiphase converters. Both blocks require the specification of
same manner, the inverse set of elementary row-switching the number of phases and their outputs are a matrix with the
transfor mations is applied to the matrix ˆ D by the matrix PT switching vector sequence and a vector with the switching
times.
 5

4. The 'Sequence' block provides the time sequence of that a multilevel modulator can be realized by means of a two-
switching vectors. level modulator.
5. The 'Trigger signals' generates the proper trigger signals The new modulation technique handles all switching states
from output levels specified in each switching vector. of the converter and it provides a sorted switching vector
6. The '5-level 5-phase cascaded full-bridge inverter' block sequence that minimizes the number of switchings. It is
is the ideal model of the multilevel voltage-source converter. obviously found that the proposed PWM offers lower THD
7. The loads are resistances with a series connected compared to conventional PWM one, thus the superiority.
inductances. Load neutral is connected to the inverter neutral.
VI. REFERENCES

V RESULTS
[1]. Levi, E., Bojoi, R., Profumo, F., Toliyat, H.A., and Williamson, S., “
Multiphase induction motor drives – a technology status review” IEE
Elec. Power appl., 2007
[2]. Singh, G.K.,” Multiphase induction motor drive research- a
survey”,Electric Power systems Research,61,pp,139-147,2002
[3]. E,E.Ward and H.Harer,”preliminary investigation of an inverter fed 5-
phase induction motor”,Proc.IEE,Vol.116, No6,pp.980-98,1969
[4]. Williamson ,S.,and Smith,A.C,”Pulsating torque and losses in
multiphase induction machines”,IEEE Trans.on industy applications
,vol.39,no.,pp,986-993,2003
[5]. J.rodriguez , J.Lai and F.Z Peng,”multi level inverters : a survey of
topologies,controls and applications,” IEEE
Trans.ind.Electron.,vol,49,no.4,pp 724-738,Aug,2002.
[6]. Q.Song, X.Zhang, F. Yu ,and C.Zhang, ”esearch on SVPWM of 5 phase
3 level inverter”,inPoc.8th ICEMS,sep.27-29,200,vol.2,pp.1418-1421.
[7]. K. Corzine and S.lu ,”comparison of hybrid propulsion drive schemes”in
Proc.IEEEElect.ship Technol symp.,jul.25-27,2005,pp.355-362.
[8]. T. J. Kim, D. W. Kang, Y. H. Lee, D. S. Hyun, ”The analysis of
conduction and switching losses in multilevel inverter system”, in proc.
of the 32th Power Electronics Specialists Conference and
Applications, pp.1363-1368, 2001.
[9]. K. A. Corzine, “Operation and design of multilevel inverters,”
University of Missouri - Rolla, Tech. Rep., 2005. [Online]. Available:
http://www.motorlab.com/
[10]. T. Bruckner and D. Holmes, “Optimal pulse-width modulation for three-
level inverters,” IEEE Trans. Power Electron., vol. 20, no. 1, pp. 82–89,
Jan. 2005.
[11]. M. Fracchia, T. Ghiara, M. Marchesoni, and M. Mazzucchelli,
“Optimized modulation techniques for the generalized N-level
converter,” in Proc. IEEE Power Electronics Specialists Conf. PESC,
vol. 2, Toledo, Spain, 29 Jun.–3 Jul. 1992, pp. 1205–1213.
[12]. C. Feng, J. Liang, and V. Agelidis, “A novel voltage balancing control
Fig : five level five phases voltages, currents
method for flying capacitor multilevel converters,” in Proc. IEEE
Industrial Electronics Society Conf. IECON, vol. 2, Roanoke, VA, 2–6
Here all five phases output voltages,currents are shown. If the Nov. 2003, pp. 1179–1184.
modulation index is high, then modulation algorithm takes [13]. J. Rodriguez, S. Bernet, B. Wu, J. Pontt, and S. Kouro, “Multilevel
advantage of all five levels of the inverter. Nevertheless, if the voltage-source-converter topologies for industrial medium-voltage
drives,”IEEE Trans. Ind. Electron., vol. 54, no. 6, pp. 2930–2945, Dec.
modulation index is low then output voltage is a three-level 2007.
waveform .with purely sinusoidal output when the modulation [14]. G. K. Singh, “Multi-phase induction machine drive research—A
index is high ,the low-order harmonics are negligible and the survey,”Electric Power Systems Research, vol. 61, no. 2, pp. 139–147,
28 Mar. 2002
total harmonic distortion (THD) is3.8%.If the modulation [15]. A. Iqbal and E. Levi, “Space vector PWM techniques for sinusoidal
index is low, then the low-order harmonics grow because of output voltage generation with a five-phase voltage source inverter,”
the three-level output and the THD increases up to 6.4%. Electric Power Components and Systems, vol. 34, no. 2, pp. 119–140,
2006.
[16]. O. López, J. Alvarez, J. Doval-Gandoy, F. D. Freijedo, A. Nogueiras,
and C. M.Peñalver, “Multilevel multiphase space vector PWM
V.CONCLUSION algorithm applied to three-phase converters,” in Proc. IEEE Industrial
Electronics Society Conf. IECON, Or-lando, FL, 10–13 Nov. 2008, pp.
A novel formulation of the modulation problem in 3288–3293.
multilevel multiphase converters has been presented in this
paper, as a part of investigation work done on the topologies
of multi level inverters. This new formulation that was carried
out in a multidimensional space, from the analysis has proved
to be simple and very useful to study the modulation process.
By using this new formulation the multilevel multiphase
SVPWM problem has been solved for converters, without
switching state redundancy. In this case, it was demonstrated