IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO.

4, JULY 2007 1453

Comparison of Traditional Inverters and

Z -Source Inverter for Fuel Cell Vehicles
Miaosen Shen, Member, IEEE, Alan Joseph, Jin Wang, Member, IEEE, Fang Z. Peng, Fellow, IEEE, and
Donald J. Adams, Member, IEEE

Abstract—In this paper, three different inverters: conventional

pulsewidth modulation (PWM) inverter, dc–dc boosted PWM in-
verter, and -source inverter were investigated and compared for
fuel cell vehicle application. Total switching device power, passive
components requirement, and constant power speed ratio of each
of these inverters were calculated. For purposes of comparison, an
example of the total switching device power, requirement of passive
components, the constant power speed ratio, and the efficiencies of
the different inverters for fuel cell vehicle powered by the same fuel
cell were conducted. The comparisons show that the -source in-
verter is very promising in applications when the boost ratio is low
(1–2).
Index Terms—DC–DC power conversion, fuel cells, inverter. Fig. 1. Typical PEM fuel cell polarization curve.

I. INTRODUCTION
F UEL cells are very promising as the emerging energy
sources in the near future. They have been used in a
variety of areas, such as domestic applications, utility applica-
tions, and traction applications [1]–[4], [11]. Unlike batteries
that have a fairly constant output voltage, the fuel cell has a
unique – characteristic and wide voltage range [15]. Fig. 1
shows the – characteristic of proton exchange membrane
(PEM) fuel cell, which is most promising for transportation
applications. This unique – curve imposes challenges on the
conditioning/interface circuits. For example, for fuel cell vehi-
cles, this results in difficulty for high-speed, and high-power
operation to achieve a great constant power speed ratio (CPSR).
In addition, because the voltage drops at high power, the
inverter has to be oversized.
Currently, there are two existing inverter topologies used
for hybrid electric and fuel cell vehicles: the conventional
three-phase pulsewidth modulation (PWM) inverter and a
three-phase PWM inverter with a dc-dc boost converter. Be-
cause of the wide voltage range and limited voltage level of
the fuel cell stack, the conventional PWM inverter topology
imposes high stresses to the switching devices and motor, and
limits the motor’s CPSR. The dc–dc boosted PWM inverter
topology can alleviate the stresses and limitations, however
suffers problems such as high cost and complexity associated
with the two-stage power conversion.
The newly proposed -source inverter [5], [6] has the unique
feature that it can boost the output voltage by introducing a
shoot through operation mode, which is forbidden in traditional
voltage source inverters. With this unique feature, the -source
inverter provides a potential cheaper, simpler, single stage ap-
proach for fuel cell vehicles. Moreover, it highly enhances the
reliability of the system because the inverter can handle mo-
mentary shoot through caused by electromagnetic interference
(EMI) without interrupting the operation.
In order to find the most suitable circuit configuration for
the power conditioner of fuel cell vehicles, this paper provides
analysis and comparisons of the three inverters for fuel cell
vehicle traction drives using total switching device power
(SDP), passive components requirement, CPSR, and efficiency,
as benchmarks.

II. SYSTEM CONFIGURATIONS FOR FUEL CELL VEHICLE

Manuscript received April 26, 2006; revised August 7, 2006. This work was
supported in part by the Department of Energy FreedomCar Program via Oak
Ridge National Laboratory and in part by the National Science Foundation under
Grant 0424039. Recommended for publication by Associate Editor T. Shimizu.
M. Shen is with Siemens VDO Electric Drives, Inc., Dearborn, MI 48120
USA (e-mail: miaosen.shen@siemens.com).
A. Joseph and F. Z. Peng are with the Department of Electrical and Computer
Engineering, Michigan State University, East Lansing, MI 48824 USA (e-mail:
josepha2@egr.msu.edu; fzpeng@egr.msu.edu).
J. Wang is with the Ford Motor Company, Dearborn, MI 48120-1261 USA
(e-mail: jwang78@ford.com).
D. J. Adams is with the Oak Ridge National Laboratory, National Transporta-
tion Research Center, Knoxville, TN 37932 USA (e-mail: adamsdj@ornl.gov).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TPEL.2007.900505
0885-8993/$25.00 © 2007 IEEE
As previously mentioned, three different inverter systems are
to be investigated: the conventional PWM inverter, the dc–dc
boosted PWM inverter, and the -source inverter. Their system
configurations for fuel cell vehicles are shown in Fig. 2 (a)–(c),
respectively.
In the traditional PWM inverter, the dc bus voltage, which
is also the output voltage of the fuel cell stack, varies with the
load. Modulation index has to be controlled to achieve the re-
quired output voltage. The boost converter in the dc–dc boosted
PWM inverter system boosts the dc voltage only when the re-
quired output voltage is higher than the fuel cell voltage. The
-source inverter outputs a required voltage by adjusting the

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 1454 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO. 4, JULY 2007

boost ratio, defined by for conventional

PWM inverter; , for dc–dc boosted PWM
inverter; and in (1) for -source inverter.
Basically, this is the ratio of the maximum dc
voltage across the inverter and minimum fuel cell
output voltage;
duty ratio of the dc–dc converter in dc–dc boosted
PWM inverter;
capacitor rms current;
switching cycle;
peak load current.

A. Total Switching Device Power Comparison

In an inverter system, each switching device has to be selected
according to the maximum voltage impressed and the peak and
average current going through it. To quantify the voltage and
current stress (or requirement) of an inverter system, switching
device power is introduced. The SDP of a switching device/cell
is expressed as the product of voltage stress and current stress.
The total SDP of an inverter system is defined as the aggregate
of SDP of all the switching devices used in the circuit. Total SDP
is a measure of the total semiconductor device requirement, thus
an important cost indicator of an inverter system. The definitions
are summarized as follows:

Fig. 2. Three inverter system configurations for fuel cell vehicles. (a) System Total Average SDP SDP
configuration using conventional PWM inverter. (b) System configuration using
dc–dc boosted PWM inverter. (c) System configuration using the Z -source
inverter. and

Total Peak SDP SDP

shoot through duty ratio with the restriction to keep the voltage
across the switches not to exceed its limit [7]. where is the number of devices used, , and
are the average and peak current through the device, respec-
tively, and is the peak voltage induced on the devices.
III. COMPARISON ITEMS, CONDITIONS, In our comparison, the input end diode in the traditional
EQUATIONS, AND RESULTS PWM inverter and the -source inverter are not considered, be-
cause it’s difficult to compare the cost of a diode and a switch
For all comparisons, the conventional PWM inverter and the of the same rating. The average and peak SDPs of conventional
dc–dc boosted PWM inverter are controlled by SPWM with PWM inverter are
third harmonic injection to achieve the maximum modulation
index possible when necessary. The -source inverter is con- SDP (1)
trolled with maximum constant boost control with modified
PWM scheme [7], [13]. To compare different inverters, some SDP (2)
parameters that are necessary for derivation are as follows:
The average and peak SDPs of dc–dc boosted PWM inverter are
maximum output power;
maximum output voltage of the fuel cell stack, SDP (3)
open circuit voltage;
PF power factor of the motor;
SDP (4)
fuel cell stack output voltage at maximum power;
modulation index of inverters, defined by the The average and peak SDPs of -source inverter are
ratio of the amplitude of the reference waveform
and that of the carrier waveform for traditional SDP (5)
SPWM;
output voltage of the boost converter in the dc–dc
boosted PWM inverter; (6)

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 SHEN et al.: COMPARISON OF TRADITIONAL INVERTERS AND -SOURCE INVERTER 1455

Fig. 3. Comparison of SDP of different inverters at different operation conditions: (a) results of peak SDP for different power factors and (b) results of average
SDP for different power factors.

The detailed derivation is in Appendix 1.

Assume that the motors for different inverters are all in con-
stant torque region, thus the motor current almost maintains con-
stant for different power. The conventional PWM inverter and
Fig. 4. Capacitor current ripple calculation.
the dc–dc boosted PWM inverter are operated with modulation
index of 1.15 with third harmonic injection or SVPWM at max-
imum power condition to minimize the required SDP. Based on
ductors are designed to limit the current ripple, and the capac-
the above equations and operation conditions, the required SDP
itors are designed based on current capacity and capacitance
of each inverter can be calculated for different load power factor
requirement.
and boost ratio, , the results are shown in Fig. 3. As mentioned
For an inductor, the size is determined by the inductance and
above, the boost ratio is defined by for conventional
the current level. The average current, , through the inductor
PWM inverter; , for dc–dc boosted PWM inverter; and
in the dc–dc boosted PWM inverter at maximum power is
in (47) for -source inverter.
For the conventional PWM inverter, it is obvious that the re- (7)
quired SDP is proportional to the apparent power, therefore the
SDP should increase linearly with the decrease of the load power For the -source inverter, the average current through the
factor. For the dc–dc boosted PWM inverter, the SDP of the in- inductor equals to that through the input end diode, based on
verter increases with the apparent power and the SDP of the power balance, the average inductor current can also be calcu-
dc–dc converter stays constant. It is similar for the -source lated by (7).
inverter, the SDP contributed by shoot through doesn’t change The inductor current ripple, , in the dc–dc boosted PWM
with the apparent power. Therefore, as load power factor re- inverter can be calculated by
duces, the SDP of conventional PWM inverter increases quicker
than the other two inverters, which is verified from Fig. 3. As (8)
can be seen from Fig. 3, if the same boost ratios are used for the
three inverters, the -source inverter requires low SDP at low where is the inductance of the inductor.
boost ratio region (1–2), especially average SDP, which is also The inductor current ripple in the -source inverter using
an indicator of thermal requirement and conversion efficiency. maximum constant boost control [7] is
When the boost ratio is above 2, the dc–dc boosted PWM in-
verter is the best in terms of SDP requirement. For very low
boost ratio (1–1.25) and high power factor, the SDP of the con- (9)
ventional PWM inverter is very similar or even smaller than the
-source inverter. where is the inductance of the inductors.
The detailed derivations of the equations are in Appendix 2. In
B. Requirement of Passive Components Comparison (8) and (9), all other parameters are known for maximum power
Passive components—inductors and capacitors are also im- operation condition, therefore the inductance requirement can
portant parts determining the inverter cost and volume. The in- be calculated by limiting the current ripple to a certain range.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 1456 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO. 4, JULY 2007

Fig. 5. PWM scheme of different inverters at a certain interval: (a) conventional PWM inverter, (b) dc–dc boosted PWM inverter, and (c) Z -source inverter.
v ; v ; v : reference signals of inverter PWM; V : reference signal of the dc–dc converter; V ; V : reference signals controlling the shoot through.

Fig. 6. Capacitor current comparison of the inverters: (a) conventional PWM inverter, (b) dc–dc boosted PWM inverter, and (c) Z -source inverter
K = (I )=(I ); I is the peak load current of the inverter, the dotted curves in (b) demonstrate when the dc–dc converter is operating with D >0, the
dotted curves in (c) demonstrates when the inverter is operating with shoot through.

For the two inductors in the -source inverter, the current where , and is the load current of phase , and .
through them and the voltage across them are exactly the same, As a result, the following equation can be derived:
therefore, they can be built on the same core with the same
size of one inductor with doubled inductance. A detailed design
process can be found in [13].
The current through the capacitor is an important factor deter-
mining the capacitor size. It is obvious that the current through
the capacitor repeats every 1/3 of the fundamental cycle. From (12)
Fig. 4, the capacitor current can be calculated by
Based on this equation, the capacitor current for different
modulation indexes and load power factors are shown in
(10)
Fig. 6(a). is defined by , where
is the output peak current.
where is the fundamental frequency. For the dc–dc boosted PWM, assuming that the control of
For conventional PWM inverter, the input current to the ca- the dc–dc converter and the inverter share the same carrier, for a
pacitor, , is assumed to be constant, which equals to the av- certain interval with modulation scheme shown in Fig. 5(b), the
erage current, . The output current of the capacitor input and output current of the capacitor can be described as
changes with the time, and it can be different for different PWM
schemes. However, the rms current keeps the same [9], [10]. For
the switching cycle with SPWM scheme shown in Fig. 5(a), the
instant current can be described as (13)

(11) (14)

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 SHEN et al.: COMPARISON OF TRADITIONAL INVERTERS AND -SOURCE INVERTER 1457

Fig. 7. Capacitor voltage ripple comparison of the inverters: (a) conventional PWM inverter, (b) dc–dc boosted PWM inverter, and (c) Z -source inverter K =
(1 ) (
Q = I :T ) ;I is the peak load current of the inverter, the dotted curves in (b) demonstrate when the dc–dc converter is operating with D > 0, the
dotted curves in (c) demonstrate when the inverter is operating with shoot through.

The resulted current depends on the relative positions of the TABLE I

reference signals. It is easier to use numerical calculation than
analytical expression. The dc–dc boosted PWM inverter oper-
ates as a conventional PWM inverter when the input voltage is
higher than the required output voltage with 0. When the
input voltage is not high enough to output the required voltage,
to minimize the switching loss, the inverter is always operated
with maximum modulation index possible, 1.15, and the dc–dc
converter operates with greater than zero to output the re-
quired voltage. Fig. 6(b) shows the calculated capacitor current
for the dc–dc boosted PMW inverter for these two conditions.
The solid curves show the capacitor current when the 0 the input voltage is high enough to output the required voltage,
with different modulation indices, the dotted curves show the
capacitor current when 0, and 1.15 for different load
power factors.
Maximum constant boost control with third harmonic injec-
tion [7] shown in Fig. 5(c) is used to control the -source in-
verter. As seen from Fig. 5(c), when the carrier is higher than
or lower than , the inverter is in shoot through state, while
in between, it operates as traditional SPWM. The input current
to the capacitor is the inductor current, which can be considered
as constant, . During traditional zero states and ac-
tive states, the output current of the capacitor is the same as the
conventional PWM inverter. During shoot through state, the two
capacitors are charging the two inductors separately through the
inverter bridge, therefore the current through the inverter bridge
is twice of the inductor current. More detailed description of the
output current of the capacitor at the interval shown in Fig. 5(c)
and (15), shown at the bottom of the page.
From this equation, we can calculate the current through the
capacitors in the -source inverter. Similarly, the -source in-
verter can be operated as a conventional PWM inverter when
SWITCHING DEVICE POWER COMPARISON EXAMPLE
and it will be operated with shoot through only when the input
voltage is low and the required output voltage is high. The re-
sulted capacitor currents for this two different operation condi-
tions are shown in Fig. 6(c).
It is note worthy that is different for different inverter
systems. For the same fuel cell stack and same power, of
the conventional PWM inverter will be higher than the other two
because of lower available voltage.
The capacitance can also be an important factor determining
its size depending on the type of the capacitor. The capacitance
is designed based on limiting the voltage ripple across the capac-
itor to limit the output current harmonic. From above analysis of
the current through the capacitor, it is easy to numerically cal-
culate the maximum voltage ripple of the capacitors under dif-
ferent conditions. The calculated results are shown in Fig. 7. The
dotted curves in (b) and (c) show the results when boost function
is needed. The result is defined by ,
where is the charge causing the maximum voltage ripple,
is the peak load current, is the switching cycle. Also,
is different for different inverter systems.

traditional zero state

(15)
shoot through state

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 1458
TABLE II
REQUIRED PASSIVE COMPONENTS
C. CPSR Comparison
CPSR is limited mainly by available dc voltage of the PWM
inverter. The fuel cell voltage decreases as the current drawn
increases, which greatly limits the motor’s power output and
efficiency at high speed. For dc–dc boosted PWM inverter and
-source inverter, the available output voltage is theoretically
infinity. To compare the CPSR limited by the inverter output
voltage, certain operation condition has to be specified. To make
a fair comparison, the same fuel cell stack and the same voltage
rating device are used by limiting the maximum output voltage
of the dc–dc converter, , the maximum voltage across the
device of -source inverter, , and the open circuit voltage of
the fuel cell, , the same, thus the maximum voltage across
the device and the boost ratio are the same for all inverters.
The modulation indexes of the conventional PWM inverter and
the dc–dc boosted PWM inverter are both 1.15 to output the
maximum voltage. The maximum obtainable output phase peak
voltage of conventional PWM inverter at peak power is
(16)
The maximum obtainable output voltage of the dc–dc boosted
PWM inverter is
(17)
For the -source inverter, the modulation index used to boost
the voltage is determined by [7]
(18)
The resulted obtainable output voltage is
(19)
Define the CPSR of conventional PWM inverter to be 1
per unit. The CPSR of dc–dc boosted PWM inverter and the
-source inverter are , and 1 2 , respectively.
D. Reliability Comparison
Compared to traditional PWM inverter and the -source
inverter, the dc–dc boosted PWM inverter uses one more active
device, which inevitably reduces the reliability of the inverter.
Also, the -source inverter can handle shoot through state,
thus momentary shoot through caused by EMI will not affect
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO. 4, JULY 2007
the inverter operation, while for traditional inverters, this might
trigger protection and cause unexpected system shut down.
IV. COMPARISON EXAMPLE
In this section, a comparison example will be conducted
based on a given fuel cell model with the following parameters:
kW
The conventional PWM inverter and the dc–dc boosted PWM
inverter are operated with modulation index of 1.15 with third
harmonic injection or SVPWM at maximum power condition.
For a fair comparison, the same voltage rating devices are used,
which means the boost ratios of the three inverters are kept the
same.
With these assumptions, the SDP of different inverters can be
calculated and shown in Table. I.
The -source inverter’s average SDP is the lowest among
the three while the conventional PWM inverter’s SDPs are the
highest in both average and peak values. Usually, the selec-
tion of the switching device is based on the rms/average current
rating, and also the average SDP is a measurement of thermal
requirement.
Table II shows the passive component comparison with the
requirements to limit the inductor current ripple to be less than
10% of its average value, and capacitor voltage ripple less than
3% of the maximum voltage, 420 V at switching frequency of
10 kHz. The -source inverter’s two inductors can be built on
one core to minimize the size and weight. In general, the re-
quired and of the -source inverter are slightly greater than
those of the dc–dc boosted PWM inverter.
For the conventional PWM inverter with the fuel cell model
described above, the fuel cell voltage is the dc voltage of the
inverter, which drops to 250 V at 200 A. From the 250 V dc,
the conventional PWM inverter can only yield 176 V to the
motor. This low motor voltage limits CPSR and lowers mechan-
ical output power and efficiency. The PWM inverter with dc-dc
boost can keep the dc voltage to 420 V, which in turn increases
CPSR by a factor of 1.68. Theoretically the -source inverter
can output whatever voltage as required. By the restriction of the
same switch voltage stress as the traditional PWM inverter and
dc–dc boosted PWM inverter, the -source inverter can increase
the CPSR by 1.34 times over the traditional PWM inverter. In
other words, the motor voltage produced by the -source in-
verters is 1.34 times that produced by the conventional PWM
inverter, thus the same motor can output 1.34 times the power
than when driven by the conventional PWM inverter.
 SHEN et al.: COMPARISON OF TRADITIONAL INVERTERS AND -SOURCE INVERTER 1459

TABLE III
OPERATION CONDITIONS AT DIFFERENT POWER

Fig. 8. Switching states sequence in one cycle.

Efficiency is an important criterion for any power converter.

High efficiency can reduce thermal requirements and cost. Fig. 9. Load voltage and current.
The inverter system is assumed to be operated in constant
torque region and space vector control is applied to all inverters.
Based on this assumption, the output voltage and current of
the inverter can be derived in Table III. From the resulted
voltage and current, the following devices are chosen for
efficiency calculation: the switches for the main inverters of
dc–dc boosted inverter and -source inverter are Powerex IPM
PM300CLA060, the switch for the dc–dc boost converter is
Powerex PM300DSA060, the switch for the traditional inverter
is Powerex IPM PM450CLA060, the input end diode of the
traditional PWM inverter and the -source inverter is IXYS
MEO 500-06DA.
For the efficiency comparison, only the semiconductor de- Fig. 10. Calculated efficiency of inverters.
vices losses are considered. For conventional PWM inverter
and dc–dc boosted PWM inverter, the IGBT loss consists of
switching loss and conduction loss, the diode loss includes
conduction loss and reverse recovery loss, the detailed calcula-
tion method can be found in [8]. For the -source inverter, the
loss of the inverter consists of the following parts: conduction
loss of the switches during non-shoot through states, switching
loss between non-shoot through states, conduction loss of the
switches during shoot through state, switching loss of the shoot
through state, free wheeling diodes conduction loss and reverse
recovery loss during non-shoot through states. The turn on
and turn off currents of the switches of shoot through state are
calculated based on the model shown in Fig. 13. The modified
PWM sequence of maximum constant boost control [13] is
implemented for efficiency comparison, in which one shoot
through period is employed in a switching cycle as shown Fig. 11. Inverter efficiency calculated using software.
in Fig. 8. The loss calculation method for traditional PWM
inverters can be found in [8] and the expression of switching
loss of each IGBT for a three-phase inverter is as follows: are turned into shoot through switching. Therefore, the modified
switching loss calculation has to be applied. There are two parts
(20) of switching losses: traditional switching (switching actions be-
tween traditional states) and shoot through switching (switching
states between shoot through state and traditional states). The
where , and are the turn on and turn off energy loss
switching loss of traditional switching can be calculated by
of the IGBT at peak current, is the switching frequency. For
the -source inverter, assume the load voltage and phase cur-
rent are as in Fig. 9, where the load current is lagging by . In
the shaded area, where the output voltage of phase is the max- (21)
imum among the three phases, all switching actions in phase

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 1460
Fig. 12. Efficiency testing and results: (a) 50-kW Z -source inverter and (b) measured efficiencies.
where and are the turn on and turn off energy
loss of the IGBT at peak current. During shoot through state,
the current from the dc side is 2 , where is the inductor
current. Assuming that the current is evenly distributed in three
phase legs, the average switching current of shoot through state
is 2 . In each cycle, there are three shoot through switching,
thus the shoot through switching loss of each IGBT is
(22)
where and are the turn on and turn off energy
loss corresponding to switching current of 2 3 .
The reverse recovery loss of the free wheeling diodes is re-
duced because some of the turn off states of the diodes are turn
into shoot through turn off, the reduction can be calculated in a
way similar as in (21).
The conduction losses the IGBTs and the diodes also change
because of the shoot through states. Assuming the shoot through
duty ratio is and the corresponding conduction losses of
IGBTs and diodes for traditional PWM inverter under the same
load current are , and , respectively, which can
be calculated based on [8], the conduction losses of the -source
inverter during traditional states become
(23)
(24)
Assuming that the inductor current is high enough so that all
IGBTs are on during shoot through state, the average current
through the IGBTs during shoot through is 2 3 , the con-
duction loss of IGBTs during shoot through is
(25)
where is the saturation voltage of the IGBT. From all
above discussion, total loss of the inverter bridge can be calcu-
lated. Also, the conduction loss and reverse recovery losses of
the input end diode of the -source inverter and the traditional
PWM inverter are considered. The inverter efficiency calcula-
tion results are shown in Fig. 10.
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO. 4, JULY 2007
From the above comparison, the -source inverter provides
the highest efficiency in most regions of the power range of the
inverter itself.
To verify the efficiency calculation of the traditional inverters,
Mitsubishi average loss simulation software [14] is used. The ef-
ficiencies of the conventional PWM inverter and dc–dc boosted
PWM inverter calculated from the software under the same op-
eration conditions are shown in Fig. 11. As can be seen from
Figs. 10 and 11, the calculated efficiencies and simulated effi-
ciencies are very close. However, because of the different op-
erating principle, the software can not be used to simulate the
efficiency of the -source inverter. To verify the efficiency cal-
culation of the -source inverter, efficiency measurement on a
50-kW -source inverter shown in Fig. 12(a) is conducted. De-
veloped for a different project, the inverter bridge is Powerex
IPM PM600CLA060 to meet high temperature requirement, the
inductor is 50 H each and built on one core, which is equiva-
lent to a 100- H inductor, also one film capacitor is connected
in parallel with the input voltage source just to minimize the
current loop. However, the loss characteristics of the two IPMs
are quite similar. The test was carried out on a RL load test and
with power factor of 0.8. The resulted efficiency is shown in
Fig. 12(b). Because of the load, the current at low power be-
comes smaller than motor current when operated in constant
torque mode, which results in higher efficiencies than the cal-
culated value at low power range. Also, because of lower power
factor is lower and losses on the passive components, the effi-
ciency at high power is slightly lower than the calculated value
as expected. Considering all these factors, the calculated effi-
ciencies of the inverters are quite accurate.
V. CONCLUSION
A comprehensive comparison of the three inverter systems
has been performed. General equations and curves of switching
device power, passive component requirement, and CPSR have
been derived for comparison. A comparison example with
detailed specifications has been carried out. For this compar-
ison example, efficiencies of different system have also been
compared and verified. The comparison results show that the
 SHEN et al.: COMPARISON OF TRADITIONAL INVERTERS AND -SOURCE INVERTER 1461

-source inverter has lower average SDP in low boost ratio b) DC–DC Boosted PWM Inverter: For the switch in the
range (1–2). Also it provides higher efficiencies in most oper- boost converter, treating the switch and the diode as a switch
ation ranges. It can increase the CPSR over the conventional cell, the maximum voltage it sustains is and the average
PWM inverter significantly. It does slightly increase the passive current through it during maximum power is
component requirement. It is note worthy that this is pure
theoretical comparison, in practical cases, for dc–dc boosted (33)
PWM inverter, the associated cost and volume increase of extra
heat sinking effort and gate drive for an extra switch is also The average switching device power of the dc–dc converter is
significant. Also, great reliability enhancement of the -source
inverter is a very important advantage. In general, the -source SDP (34)
inverter is very competitive in low boost ratio range (1–2), in
which most fuel cells reside. In cases when a low voltage fuel Suppose the current through the inductor in the boost con-
cell is used, and boost ratio much higher than 2 is needed, the verter is constant, the peak current through the switch is the
dc–dc boosted PWM inverter is the best configuration. In this same as the average current. The peak switching device power
paper, the comparison is limited to a fuel cell vehicle without is
battery, all configurations can be modified to fuel cell-battery
hybrid vehicles by adding a battery [12] to incorporate the SDP (35)
regenerative braking function without significant change of the
main circuit. The voltage stress of the inverter switches is . The RMS
phase voltage at modulation index of M is
APPENDIX
1) Switching Device Power Derivation: (36)
a) Traditional PWM Inverter: For the traditional PWM
inverter, the output phase RMS voltage at peak power is The RMS line current is

(26) (37)

With motor power factor of , the output line RMS current The average current through switches under maximum power is
is
(38)
(27)
The average switching device power of the system is
Because the line current is evenly shared by two switches in
a line cycle, the average current through each switch is SDP SDP
(39)

The peak switch current of the inverter is

(28)
(40)
The maximum voltage stress of the switches occurs when the
output power is zero, and the fuel cell voltage reaches its max- The peak switching device power of the system is
imum value, which is
SDP SDP
(29)
(41)
The total average switching device power of the circuit is
c) -Source Inverter: For the source inverter, the cur-
(30) rent through the inverter switches consists of two elements, one
is the current to the load and the other is the current through
The peak current through the switches is the peak line current them during the shoot through state. Because of the symmet-
rical structure of the inverter, the current during shoot through
(31) in terms of average is evenly distributed in three parallel paths.
The current through the inverter during shoot through is twice
The total peak switching device power of the traditional PWM of the inductor current. Therefore, the average current value in
inverter is shoot through period through each switch is

(32) (42)

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 1462 IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 22, NO. 4, JULY 2007

Based on this model, we can have

(49)
(50)

From which we have

(51)
Fig. 13. Inverter model during shoot through.
The peak current through the switch occurs when the line
current of phase A is at its peak, which is
where is the inductor current. From the input end, the average
current through the diode is equal to the sum of the average (52)
current through inductor and capacitor . In steady state,
the average current through the capacitor is zero, the average Another possible condition is that the peak current through
current through the inductor equals to that of the diode. The the switches occurs during the active states, then the current
output power of the fuel cell stack under maximum power is shown in (52) is the peak current.
, therefore, the average current through the diode as well as The peak switching device power of the inverter is
the inductor is:
SDP
(43)

While in active states, the average current is the same as a (53)

conventional PWM inverter, therefore the overall average cur-
rent through inverter switches is 2) Passive Component Requirement:
a) DC–DC Boosted PWM Inverter: For the inductor in
the boost converter, the current ripple can be calculated by the
current increase when the switch is on

(44) (54)

where is the shoot through period in a switching cycle where is the duty cycle.
is the RMS output phase voltage. With the control method pre- b) -Source Inverter: When the inverter is in a shoot
sented in [7], , and can be expressed as through state, the voltage across the inductor is the voltage
across the capacitor. Therefore, the current ripple of the in-
(45) ductor can be calculated as

(55)
(46)
where is the voltage across the capacitor
Voltage stress of the inverter switches is
(56)
(47)
We have
The average switching device power of the inverter is
(57)

(48)
To calculate the peak current through the switches during
shoot through, we assume that when the switches are on they
are pure resistors with the same resistance, which is shown in
Fig. 13.
REFERENCES
[1] G. A. O’Sullivan, “Fuel cell inverters for utility applications,” in Proc.
IEEE PESC’00, 2002, pp. 1191–1194.
[2] A. M. Tuckey and J. N. Krase, “A low-cost inverter for domestic fuel
cell applications,” in Proc. IEEE PESC’02, 2002, pp. 339–346.
[3] P. T. Krein and R. Balog, “Low cost inverter suitable for medium-
power fuel cell sources,” in Proc. IEEE PESC’02, 2002, pp. 321–326.
[4] J. Mazumdar, I. Batarseh, N. Kutkut, and O. Demirci, “High frequency
low cost DC-AC inverter design with fuel cell source for home appli-
cations,” in Proc. IEEE IAS’02, 2002, pp. 789–794.

Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
 SHEN et al.: COMPARISON OF TRADITIONAL INVERTERS AND -SOURCE INVERTER
[5] F. Z. Peng, “Z -source inverter,” IEEE Trans. Ind. Appl., vol. 39, no. 2,
pp. 504–510, Mar./Apr. 2003.
[6] F. Z. Peng, A. Joseph, J. Wang, M. Shen, L. Chen, Z. Pan, E.
Oritz-Rivera, and Y. Huang, “Z -source inverter for motor drives,”
IEEE Trans. Power Electron., vol. 20, no. 4, pp. 857–863, Jul. 2005.
[7] M. Shen, J. Wang, A. Joseph, F. Z. Peng, L. M. Tolbert, and D. J.
Adams, “Constant boost control of the Z -source inverter to minimize
current ripple and voltage stress,” IEEE Trans. Ind. Appl., vol. 42, no.
3, pp. 770–778, May/Jun. 2006.
[8] Powerex, “General Considerations: IGBT & IPM Modules,” Tech.
Rep., Appl. notes A10-A27, 2007.
[9] J. W. Kolar, H. Ertl, and F. C. Zach, “Calculation of the passive and
active component stress of three phase PWM converter systems with
high pulse rate,” in Proc. EPE Conf. Rec., Aachen, Germany, 1989, pp.
1303–1311.
[10] A. M. Hava, R. J. Kerkman, and T. A. Lipo, “Simple analytical and
graphical methods for carrier based PWM-VSI drives,” IEEE Trans.
Power Electron., vol. 14, no. 1, pp. 49–61, Jan. 1999.
[11] K. Rajashekara, “Power conversion and control strategies for fuel cell
vehicles,” in Proc. IEEE IECON’03, 2003, pp. 2865–2870.
[12] F. Peng, M. Shen, and K. Holland, “Application of Z -source inverter
control for for traction drive of fuel cell—battery hybrid electric ve-
hicles,” IEEE Trans. Power Electron., vol. 22, no. 3, pp. 1054–1061,
May 2007, in Proc. IEEE Ind. Appl. Soc. Annu. Meeting, Hong Kong,
China, 2005, pp. 1651–1656.
[13] M. Shen, A. Joseph, Y. Huang, F. Z. Peng, and Z. Qian, “Design and
development of a 50 kW Z -source inverter for fuel cell vehicles,” in
Proc. Int. Conf. Power Electron. Mot. Contr., Shanghai, China, 2006,
pp. 1076–1080.
[14] Powerex, “Global Power-Semiconductor Solution Provider,” 2007
[Online]. Available: http://www.pwrx.com/
[15] C. Wang, M. Nehrir, and S. Shaw, “Dynamic models and model vali-
dation of PWM fuel cells using electrical circuits,” IEEE Trans. Power
Conv., vol. 20, no. 2, pp. 442–451, Jun. 2005.
Miaosen Shen (S’04–M’07) received the B.S. and
M.S. degrees in electrical engineering from Zhejiang
University, Hangzhou, China, in 2000 and 2003,
respectively, and the Ph.D. degree in electrical
engineering from Michigan State University, East
Lansing, in 2007.
He joined Siemens VDO Electric Drives, Inc.,
Dearborn, MI, in 2007 working on hybrid electric
drives. His research interests include motor drives,
power factor correction technique, and electronic
ballast for HID lamps.
Alan Joseph received the B.S.E.E. degree from Oak-
land University, Rochester, MI, in 1998 and the M.S.
degree in electrical engineering from Michigan State,
East Lansing, in 2002.
He is with the Department of Electrical and
Computer Engineering, Michigan State University,
where he conducts research in the areas of multilevel
converters, power conversion for alternative energy
sources, and variable frequency drives.
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY WARANGAL. Downloaded on February 27,2022 at 08:06:40 UTC from IEEE Xplore. Restrictions apply.
1463
Jin Wang (S’01–M’05) received the B.S. degree
from Xi’an Jiaotong University, Xi’an, China, in
1998, the M.S. degree from Wuhan University,
Wuhan, China, in 2001, and the Ph.D. degree from
Michigan State University, East Lansing, in 2005,
all in electrical engineering.
He joined Ford Motor Company, Dearborn, MI,
in 2005. He is currently a Core Engineer for power
electronics systems in Ford’s hybrid vehicles. His
research interests include hybrid electric vehicles
(HEVs)/fuel cell vehicles (FCVs), multilevel con-
verters, dc-dc converters, flexible ac transmission systems (FACTs) devices,
and DSP-based control systems.
Fang Z. Peng (M’92–SM’96–F’05) received the B.S.
degree in electrical engineering from Wuhan Univer-
sity, Wuhan, China, in 1983 and the M.S. and Ph.D.
degrees in electrical engineering from Nagaoka Uni-
versity of Technology, Nagaoka, Japan, in 1987 and
1990, respectively.
He was with Toyo Electric Manufacturing Com-
pany, Ltd., from 1990 to 1992 as a Research Scien-
tist, and was engaged in research and development of
active power filters, flexible ac transmission systems
(FACTS) applications, and motor drives. From 1992
to 1994, he worked with the Tokyo Institute of Technology, Tokyo, Japan, as a
Research Assistant Professor, initiated a multilevel inverter program for FACTS
applications and a speed-sensorless vector control project. From 1994 to 2000,
he was with Oak Ridge National Laboratory (ORNL), Knoxville, as a Research
Assistant Professor, at University of Tennessee, Knoxville, from 1994 to 1997,
and was a Staff Member, Lead (Principal) Scientist of the Power Electronics
and Electric Machinery Research Center, ORNL, from 1997 to 2000. In 2000,
he joined Michigan State University, East Lansing, as an Associate Professor
and is now a Professor with the Department of Electrical and Computer Engi-
neering. He holds over 10 patents.
Dr. Peng received the 1996 First Prize Paper Award and the 1995 Second Prize
Paper Award of the Industrial Power Converter Committee at the IEEE/IAS An-
nual Meeting; the 1996 Advanced Technology Award of the Inventors Clubs of
America, Inc.; the International Hall of Fame; the 1991 First Prize Paper Award
in the IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS; the 1990 Best Paper
Award in the Transactions of the IEE of Japan; and the Promotion Award of the
Electrical Academy. He was Chair of Technical Committee for Rectifiers and
Inverters of IEEE Power Electronics Society from 2001 to 2005 and was an As-
sociate Editor for the IEEE TRANSACTIONS ON POWER ELECTRONICS from 1997
to 2001 and has been an Associate Editor again since 2005.
Donald J. Adams (M’95) received the B.S. degree
from the University of Mississippi, Oxford, in
1973 and the M.S. degree from the University of
Tennessee, Knoxville, in 1977, both in mechanical
engineering.
He is the Director of the Power Electronics and
Electric Machinery Research Center, Oak Ridge Na-
tional Laboratory (ORNL), Knoxville, where he has
been employed for 29 years. He is on the Governing
Board of the NSF Center for Power Electronics Sys-
tems (which consists of five universities and over 80
industrial partners). He is the holder of seven patents with two pending, and
has authored numerous publications. His research interests include advanced
inverters and adjustable-speed drives, power transmission and distribution re-
search and development, electric machines, and power quality, efficiency, and
power measurements.
Mr. Adams is a Registered Professional Engineer in the State of Tennessee.