3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

A New Control Strategy for a Class of Multiple-Input

DC-DC Converters

V. A. K. Prabhala Mostafa Khazraei Mehdi Ferdowsi

Electrical Engineering Dept. Electrical Engineering Dept. Electrical Engineering Dept.
Missouri S&T Missouri S&T Missouri S&T
Rolla, USA Rolla, USA Rolla, USA
[email protected] [email protected] [email protected]

Abstract—This paper proposes a new control scheme for a class loops for controlling the flow of energy from each source.
of multi-input dc-dc power converters. The new control method However, these methods have proven to show slow dynamics
is based on coupling the signals from independent control loops and poor performance in case of sudden disturbances in the
of conventional controller for a multi-port converter. In this system. There have been some efforts in the literature to
paper, the non-restricted double input buck converter is overcome this issue with the conventional controllers;
introduced. The properties of this converter and its steady-state however, few proposed control schemes have high degree of
equations along with its small-signal model are reviewed. Next, complexity which makes them less appealing for
the conventional control method is reviewed and the new control implementation [13-22]. This paper proposes a new control
method is introduced based on coupling the voltage and current
method based on the coupling of current and voltage control
loops from the conventional controller. Then, the small-signal
loops of a conventional controller to improve the performance
model of the converter with the new control method is analyzed
and the converter transfer functions are derived. Finally, the of the system.
converter transfer functions are used to find the output In this paper, in Section II the non-restricted double input
impedance of the converter in order to compare the performance (DI) buck converter is introduced and the steady-state and
of the new controller with the conventional controllers. dynamic equations describing the behavior of the converter are
Simulation results are provided to verify the results predicted by provided. In Section III the new control method is explained
the small signal analysis. and the performance of the new control method is analyzed and
compared with the conventional method using the small-signal
I. INTRODUCTION model. In Section IV simulation results are provided to verify
the results predicted by small-signal analysis.
Renewable energy sources are fast becoming the primary
means for overcoming the worldwide energy crisis which has
resulted in extensive research and development in the area of II. NON-RESCTRICTED DOUBLE-INPUT BUCK CONVERTER
power conversion techniques. Specifically, multi-input The converter under study in this paper is a non-restricted
converters have been in the center of attention of engineers in double-input buck converter introduced in [3]. The circuit
academia and industry because of their high flexibility for diagram of this converter is illustrated in Fig. 1. As pictured,
combining energy from various sources. The multi-input this converter employs two active switches S 1 and S 2 and two
converters can easily combine the energy from two or more diodes D 1 and D 2 . The modes of operation of this converter are
sources with intermittent nature and transfer it to a single load listed in TABLE I. According to TABLE I there is no mode
or a dc bus. Numerous publications propose using a multi-input restriction for the operation of this converter meaning that all 4
converter to combine various energy sources instead of the
possible modes of operation resulting from independently
traditional method of interfacing each source to the dc bus with
turning the switches ON and OFF can be used in the operation.
a separate conventional single-input converter [1-12].
Therefore, both switches S 1 and S 2 can be ON at the same time
Additionally, the multi-input converters are gaining (Mode IV) and thus the inductor can be energized by both the
attention in the field of hybrid electric vehicles as well [6, 7, sources simultaneously. The steady-state output voltage of this
14, 18]. In hybrid electric vehicles, by using a multi-input converter is found in [18] by averaging the output voltage over
converter, the advantages of a source with high energy density one switching period,
such as a battery can be combined with the advantages of a
source with high power density such as an ultra-capacitor. This
way the battery can supply constant current throughout the V0 D1V1  D2V2 (1)
operation and the ultra-capacitor can cater the peak load current
when necessary. where D 1 and D 2 are the duty cycles of switches S 1 and S 2 in
Conventional methods of controlling a multi-input the steady-state condition, V 1 and V 2 are the voltages of two
converter make use of separate current and voltage control sources feeding the converter.

86*RYHUQPHQWZRUNQRWSURWHFWHGE\86FRS\ULJKW

ICRERA 2014 508

3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

iS1 S1 + VL -
TABLE I. MODES OF OPERATION OF A NON-RESTRICTED DOUBLE-
L INPUT BUCK CONVERTER
iL
V1 D1 Mode S1 S2 VL V D1 V D2
R +
C Vo I On Off V 1 ‫ޤ‬V o ‫ޤ‬V 1 0
_
II Off On V 2 ‫ޤ‬V o 0 ‫ޤ‬V 2
V2 D2
III Off Off ‫ޤ‬V o 0 0
iS2 S2
IV On On V 1 +V 2 ‫ޤ‬V o ‫ޤ‬V 1 ‫ޤ‬V 2
Figure 1. Circuit diagram of the non-restricted DI buck converter [3].

The equations describing the dynamic behavior of this 1

converter can be found by averaging the inductor voltage and D2V2 (  sC )
iˆs 2 ( s ) R (9)
capacitor current equations over one switching period [3]. Gis 2 d 2 ( s ) IL 
dˆ ( s )
2 dˆ1 ( s ) vˆ1 ( s ) vˆ2 ( s ) 0
L
s 2 LC  s  1
The average source currents can be found by this method as R
well,
III. COUPLING CONTROL STRATEGY
d iL (t )
v L (t ) L
TS
d1 (t ) v1 (t )  d 2 (t ) v2 (t )  v0 (t ) (2) Fig. 2 shows the block diagram of the conventional control
TS TS TS TS
dt method for a DI buck converter. According to the control
strategy, one of the input currents is controlled keeping the
d v0 (t ) v0 (t ) output voltage constant. Fig. 2 shows the conventional control
iC (t ) C
TS
i L (t ) 
TS (3)
TS
dt TS
R method where a current compensator is used to control one of
the input currents and a voltage compensator is utilized to
control the output voltage. It can be seen that the control
iS 1 (t ) TS
d1 (t ) iL (t ) TS (4) signal from the voltage compensator is used to generate d 2 , the
switch command for switch S 2 . It can also be observed that
iS 2 (t ) d 2 (t ) iL (t ) (5) control signal from the current compensator is used to
TS TS
generate d 1 , the switch command for switch S 1 .
The converter transfer functions are needed for the small- In many applications, the fast dynamics of i s1 is desirable.
signal analysis of the effect of the new controller. The Hence, a new modified method is introduced called the
converter transfer functions can be found by linearizing (2)-(5) coupling method which can help improve the dynamic
about an operating point and getting the Laplace transform of response of the converter even with large load variations. As
the resulted equations [3]. The required transfer functions for shown in Fig. 3, this method is based on coupling the output
the rest of the analysis are the transfer functions of the output of the voltage and current compensators. This coupling makes
voltage and switch currents to the duty cycles, both the compensators to participate in the regulation of the
input current and output voltage. When i ref1 is decreased the
output of current compensator is decreased and hence d 1 is
vˆ0 ( s ) V1 decreased and it reduces i s1 . Moreover, based on a negative
Gv 0 d 1 ( s ) (6) sign in the coupling method, the output of current
dˆ ( s )
1 dˆ 2 ( s ) vˆ1 ( s ) vˆ2 ( s ) 0 s LC  s
2 L
1 compensator will make d 2 to increase faster and hence i s2
R increases faster. Since the load is constant, the increase in i s2
helps to decrease i s1 faster. The similar phenomenon happens
vˆ0 ( s ) V2 for the voltage loop as well.
Gv 0 d 2 ( s )
dˆ ( s )
2 dˆ1 ( s ) vˆ1 ( s ) vˆ2 ( s ) 0 s 2 LC  s
L
1 (7) The small signal model developed in the previous section
R is used to design the current and voltage compensators. In the
conventional method, the current and the voltage loops are
1 considered to be independent and the voltage and current
 sC )
D1V1 ( compensators are designed accordingly. Figs. 4 and 5 show
iˆs1 ( s ) R
Gis1d 1 ( s ) IL  (8) the voltage and current loops of the converter. G vc (s) and
dˆ ( s )
1 dˆ 2 ( s ) vˆ1 ( s ) vˆ2 ( s ) 0
L
s 2 LC  s  1 G ic (s) are designed to get desired phase and gain margins for
R the loop gains T v (s) and T i (s). Transfer functions G v0d2 (s) and
G is1d1 (s) are derived in the previous section from the small
signal model of the DI buck converter.

ICRERA 2014 509

3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

is1
is1
S1
S1 V1
V1 Double-Input L +
o V0
Double-Input L + DC-DC
o V0 Converter a
DC-DC
a d -
Converter V2
d -
V2 is2
is2
S2
S2

S1 S2
S1 S2 S Voltage
PWM1 + PWM22 + V Vref
Current VC,D1 S VC,D2 Voltage C,D2
Compensator
PWM1 PWM22 Compensator Vref
Iref1 Compensator
Current
-1
Iref1 Compensator VC,D1

Figure 2. Conventional control method. Figure 3. Coupling control method.

v̂ref d̂ 2 iˆref 1 d̂1

+
- Gvc (s) Gv 0 d 2 ( s ) +
- Gic (s) Gis1d 1 ( s )
v̂0 iˆs1
Tv (s ) Ti (s )

Figure 4. Voltage loop of the DI buck converter. Figure 5. Current loop of the DI buck converter.

80 80

60 60

40 40

20 20
20 ˜ log  T v ( j ˜ Z )  20 ˜ log  T i( j ˜ Z ) 
0 0
20 ˜ log  Gv0d2 ( j ˜ Z )  20 ˜ log  Gis1d1 ( j ˜ Z ) 
 20  20

 40  40

 60  60

 80  80
3 4 5 6 7 3 4 5 6 7
10 100 1u10 1u10 1u10 1u10 1u10 10 100 1u10 1u10 1u10 1u10 1u10
Z Z
2 ˜S 2 ˜S

90 90

0 0
AngT v ( Z ) AngT i( Z )  90
 90

AngGv0d2 ( Z )  180 AngGis1d1( Z )   180

 270  270

 360  360
3 4 5 6 7 3 4 5 6 7
10 100 1u10 1u10 1u10 1u10 1u10 10 100 1u10 1u10 1u10 1u10 1u10
Z Z
( 2 ˜S ) ( 2 ˜S )

Figure 6. Bode magnitude and phase plots of G v 0 d 2 ( s ) and T (s ) . Figure 7. Bode magnitude and phase plots of Gis1d 1 ( s ) and Ti (s) .
v

The voltage compensator is Type III and the current § s ·§ s ·

compensator is Type II. The designed compensators given in 100¨¨1  ¸¸¨¨1  ¸¸
equations (12) and (13) are used for the proposed method too. © 2 S 1000  ¹© 2S 1000  ¹ (12)
G vc ( s)
From Figs. 6 and 7, it can be observed that the voltage and § s ·§ s ·
s¨¨1  ¸¸¨¨1  ¸¸
current loop gains T v (s) and T i (s) have improved phase and © 2S 50000  ¹© 2 S 50000  ¹
gain margins compared to open loop gains G v 0 d 2 ( s ) and
Gis1d 1 ( s ). § s ·
650¨¨1  ¸¸
Tv ( s ) Gvc ( s )Gv 0 d 2 ( s ) (10) © 6S 500  ¹ (13)
Gic ( s )
§ s ·
s¨¨1  ¸
Ti ( s ) Gic ( s )Gis1d 1 ( s ) (11) © 6S 8000 ¸¹

ICRERA 2014 510

3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

80

60
š š
iˆs1 d 1 V1  d 2 V2 L
40

20
20 ˜ log  Gv0is1_coup ( j ˜ Z ) 
iˆL 0
š š š + š 20 ˜ log  Gv0is1_conv ( j ˜ Z ) 
v1 I L d1 D1 i L Dv  20
_ 1 1
 40

+  60
š
iˆs 2 C R v0  80
3 4 5 6 7
10 100 1u10 1u10 1u10 1u10 1u10
_
Z
2 ˜S
š š š + š
90
v2 I L d2 D2 i L D v
_ 2 2
0
AngGv0is1_coup ( Z )  90

AngGv0is1_conv ( Z )  180
 270
š š
d1 + d2 + Gvc(s) š  360
v ref 10 100 1u10
3
1u10
4
1u10
5
1u10
6 7
1u10
š Gic(s) -1 Z

i ref 1 ( 2 ˜S )

Figure 8. Small signal model for DI buck converter with coupling Figure 9. Bode magnitude and phase plots of Gv 0is1 ( s ) for both the
control. methods.

80

60
Gv 0is1 _ conv ( s ) Gic ( s )Gv 0 d 1 ( s ) (14)
40

20
20 ˜ log  Z0_coup ( j ˜ Z )  Gic ( s )Gv 0 d 1 ( s )
20 ˜ log  Z0_conv ( j ˜ Z ) 
0 Gv 0is1 _ coup ( s )
 20
(1  Gv 0 d 1 ( s )Gvc ( s )) (15)
 40

 60 Fig. 9 shows the bode plots of Gv 0is1 ( s ) transfer function

 80
3 4 5 6 7
for both the conventional and coupling control methods. As
10 100 1u10 1u10 1u10 1u10 1u10
Z
seen the gain of the transfer function for the new control
2 ˜S method is less than the gain for the conventional method. It
90
shows that the sensitivity of the output voltage to changes in
0 i s1 for the new method is less than the conventional method.
AngZ0_co up ( Z )  90 Hence, the dynamic response of the output voltage due to
AngZ0_co nv ( Z )  180 change in input current is faster for the proposed control
 270 method.
 360
10 100 1u10
3
1u10
4
1u10
5
1u10
6
1u10
7
To get the output impedance transfer function an ac current
Z
( 2 ˜S )
source is added to the circuit in Fig. 8. Again, by doing some
mathematical manipulation the Z 0 ( s) vˆ0 / iˆ0 for both
Figure 10. Bode magnitude and phase plot of Z 0 (s) for both the methods. conventional and coupling method can be obtained and is given
by (16) and (17).
Fig. 8 shows the small signal model for the coupling
method. The model is used to derive Gv 0is1 ( s ) vˆ0 / iˆs1 and
output impedance Z 0 (s). By doing some mathematical Gic ( s ) D1V1
sL 
manipulation the Gv 0is1 ( s ) for both conventional and coupling (1  Gic ( s ) I L ) (16)
Z 0 _ conv
sL(1  sRC ) G ( s) D1V1 (1  sRC )
methods can be obtained as given by (14) and (15).  1  V2Gvc ( s )  ic
R R(1  Gic ( s) I L )

R
[ sL(1  I L Gic ( s ))  (V2  V1 )Gic ( s ) D1 ]
(1  sRC ) (17)
Z 0 _ coup ( s )
R R
(1  I L Gic ( s )){1  Gvc ( s )(V1  V2 )}  sL(1  I L Gic ( s ))  (V2  V1 )Gic ( s ) D1  (V2  V1 )Gic ( s ) I L Gvc ( s )
(1  sRC ) (1  sRC )

ICRERA 2014 511

3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

Fig. 10 shows the output impedance Z 0 (s) transfer function 56

V0 with coupling method

for the converter with both the control methods. As seen the 55

output impedance for the new method is less than the 54

conventional one.

V (V)
53

It shows that the sensitivity of the output voltage to change

0
in the load for the new control method is less than the 52 V with conventional method
0

conventional method. Hence, the output voltage disturbance 51

due to changes in load current is rejected sooner for the new 50

method compared to the conventional method. Simulation 0.009 0.01 0.011 0.012
t (s)
0.013 0.014 0.015 0.016

results which are presented in the following section Figure 11. Output voltage v 0 waveform for both conventional and coupling
completely validate the small signal analysis. methods for a step change in i ref1 from 2 A to 1 A.

IV. SIMULATION RESULTS 2.5

A DI buck converter with V 1 = 75 V, V 2 = 60 V, f = 50

kHz, V 0 = 54 V, L = 100 —H, C = 50 —F is simulated with both 2
I with coupling method
s1

the control methods. At t = 0.01 s, there is a step change in i ref1 I with conventional method
s1

from 2 A to 1 A. Fig. 11 shows the output voltage v 0 for both

I (A)
1.5

s1
the conventional and coupling methods. As predicted based on
the small signal analysis, the gain of Gv 0is1 ( s ) with coupling 1

method (as shown in Fig. 9) is less than that with conventional

method, thus the output voltage disturbance when current 0.5
0.009 0.01 0.011 0.012 0.013 0.014 0.015 0.016

reference i ref1 changes is smaller in the case of coupling t (s)

method. Figs. 12 and 13 show the results for i s1 and i s2 , Figure 12. Average current from input source 1 i s1 for both conventional and
respectively. Both the figures clearly show that the coupling coupling methods for step change in i ref1 from 2 A to 1 A.
method has a better dynamic performance compared to the
conventional method. As explained earlier, in the coupling 2.5

method, the output of current compensator will make d 2 to

increase faster, thus i s2 increases faster. With the load 2
remaining constant, the increase in i s2 helps to decrease i s1
faster. Again at t = 0.01 s step change in ORDGIURPȍWR
I (A)

ȍ LV LQWURGXFHG )LJ4 shows output voltage v 0 for the both 1.5
I with coupling method
s2

s2
conventional and coupling methods. As predicted based on the
small signal model, it can be seen from Fig. 10 that Z 0 (s) with 1
I with conventional method
s2
coupling method is less compared to that with conventional
method. Thus the output voltage disturbance for the load 0.5
0.009 0.01 0.011 0.012 0.013 0.014 0.015 0.016
changes are less for coupling method compared to the t (s)
conventional method. Figs. 15 and 16 show the results for i s1 Figure 13. Average current from input source 2 i s2 for both conventional and
and i s2 currents respectively. It can be seen that the coupling coupling methods for step change in i ref1 from 2 A to 1 A.
method shows better dynamic performance compared to the
conventional method. 54.5

V with coupling method

0

V. CONCLUSION
54
In this paper a new control method for controlling a non-
V (V)

restricted double input buck converter has been introduced. A

0

small-signal analysis of the converter with the proposed

53.5 V with conventional method
controller was provided to compare the performance of the 0

proposed controller with the conventional controller. The

small-signal analysis suggested that the converter with the 53
proposed controller shows better output impedance 0.009 0.01 0.011
t (s)
0.012 0.013 0.014 0.015

characteristics and has better output voltage sensitivity to Figure 14. Output voltage v 0 waveform for both conventional and coupling
changes in current reference. Thus it has better dynamic PHWKRGVIRUVWHSFKDQJHLQORDGIURPȍWRȍ
performance in comparison to the conventional controller.
Simulation results were provided to verify the analytical results
from the small-signal analysis.

ICRERA 2014 512

3rd International Conference on Renewable Energy Research and Applications Milwakuee, USA 19-22 Oct 2014

1.5
IEEE Power Electronics Specialists Conf. (PESC), Jun. 2002, vol. 4, pp.
1.4 1685–1690.
Is1 with coupling method
1.3
[8] H. Li, Z. Du, K. Wang, L. M. Tolbert, and D. Liu, “A hybrid energy
Is1 with conventional method system using cascaded H-bridge converter,” in Proc. IEEE Industrial
1.2
Applications Conf., vol. 1, Oct. 2006, pp. 198-203.
I (A)

1.1 [9] Y. M. Chen, Y. C. Liu, and F. Y. Wu, “Multi-input converter with

s1

1 power factor correction, maximum power point tracking, and ripple-free

input currents,” IEEE Trans. Power Electronics, vol. 19, pp. 631-639,
0.9
May 2004.
0.8
[10] Y. M. Chen, Y. C. Liu, F. Y. Wu, and Y. E. Wu, “Multi-input converter
0.009 0.01 0.011 0.012 0.013 0.014 0.015 with power factor correction and maximum power point tracking
t (s)
features,” in Proc. IEEE Applied Power Electronics Conf. and
Figure 15. Average current from input source 1 i s1 for both conventional and Exposition (APEC), vol. 1, Mar. 2002, pp. 490-496.
FRXSOLQJPHWKRGVIRUVWHSFKDQJHLQORDGIURPȍWRȍ [11] Y. M. Chen, Y. C. Liu, and F. Y. Wu, “Multi-input DC/DC converter
based on the multiwinding transformer for renewable energy
applications,” IEEE Trans. Industry Applications., vol. 38, pp. 1096-
4
1104, Aug. 2002.
3.5
[12] B. G. Dobbs and P. L. Chapman, “A multiple-input dc-dc converter
topology,” IEEE Power Electronics Letters, vol. 1, pp. 6-9, Mar. 2003.
3
I with conventional method
s2 [13] N. D. Benavides and P. L. Chapman, “Power budgeting of a multiple-
I (A)

Is2 with coupling method input buck-boost converter,” in Proc. IEEE Trans. Power Electronics,
vol. 20, pp. 1303-1309, Nov. 2005.
s2

2.5

[14] K. P. Yalamanchili and M. Ferdowsi, “Review of multiple input DC-DC

2 converters for electric and hybrid vehicles,” in Proc. IEEE Vehicle
Power and Propulsion Conf., 7-9 Sept. 2005, pp. 160-163.
1.5
0.009 0.01 0.011 0.012 0.013 0.014 0.015
[15] S. H. Choung and A. Kwasinki, “Multi-input dc-dc converter topologies
t (s) comparison,” in Proc. IEEE Industrial Electronics Conf. (IECON), Nov.
Figure 16. Average current from input source 2 i s2 for both conventional and 2008, pp. 2359-2364.
FRXSOLQJPHWKRGVIRUVWHSFKDQJHLQORDGIURPȍWRȍ [16] Y. Li, Y. Dongsheng, and R. Xinbo, “Interleaved dual-edge modulation
scheme for double-input converter to minimize inductor current ripple,”
in Proc. IEEE Power Electronics Specialists Conference (PESC), 15-19
REFERENCES Jun. 2008, pp. 1783-1789.
[17] N. D. Benavides and P. L. Chapman, “Parametric average-value
[1] H. Tao, A. Kotsopoulos, J. L. Duarte, and M. A. M. Hendrix, “Family of modeling of multiple-input buck converters,” in Proc. Canadian
multiport bidirectional dc-dc converters,” in Proc. IEEE Electric Power Conference on ECE, 22-26 Apr. 2007, pp. 990 – 993.
Applications, vol. 153, May 2006, pp. 451 - 458.
[18] K. P. Yalamanchili, M. Ferdowsi, and K. Corzine, “New double input
[2] Y. C. Liu and Y. M. Chen, “A systematic approach to synthesizing dc-dc converter for automotive applications,” in Proc. IEEE Vehicle
multi-input dc/dc converters,” in Proc. IEEE Power Electronics Power and Propulsion Conf. (VPPC), 6-8 Sept. 2006, pp 1-6.
Specialists Conf. (PESC), Jun. 2007, pp. 2626-2632.
[19] P. J. Villegas, J. Sebastian, M. M. Hernando and F. Nuno, “Average
[3] V. A. K. Prabhala, “Control and applications of double input dc-dc current mode control of two-input buck post-regulators used in power
power electronic converters ,” M. S. Thesis, Missouri University of factor correctors,” IEEE Trans. Industrial Electronics, vol. 6, pp. 569-
Science & Technology, Rolla, May 2010. 576, Jun 1999.
[4] K. Gummi and M. Ferdowsi, “Synthesis of double-input dc-dc [20] D. Somayajula and M. Ferdowsi, “Power sharing in double-input
converters using single pole triple throw switch as a building block,” in buckboost converters using offset time control,” in Proc. IEEE 24th
Proc. IEEE Power Electronics Specialists Conf. (PESC), Greece, Jun. Applied Power Electronics Conference and Exposition (APEC),
2008. Washington D.C., Feb. 2009, pp. 1091-1096.
[5] V. A. Prabhala, M. Cespedes and Jian Sun, “Implementation of DQ
[21] V. A. K. Prabhala, D. Somayajula and M. Ferdowsi, “Power sharing in a
domain control in DSP and FPGA,” in Proc. IEEE 27th Applied Power
double-input buck converter using dead-time control”, in Proc. IEEE
Electronics Conference and Exposition (APEC), Orlando, Feb. 2012, pp.
Energy Conversion Congress and Exposition (ECCE), San Jose, Sep.
1439 - 1444.
2009, pp. 2621 - 2626.
[6] R. M. Schupbach, J. C. Balda, M. Zolot, and B. Kramer, “Design
[22] M. Khazraei, V. A. K. Prabhala, R. Ahmadi, and M. Ferdowsi, “Solid-
methodology of a combined battery-ultracapacitor energy storage unit
state transformer stability and control considerations”, in Proc. IEEE
for vehicle power management,” in Proc. IEEE Power Electronics
29th Applied Power Electronics Conference and Exposition (APEC), Fort
Conf., 15-19 Jun. 2003, vol. 1, pp. 88-93.
Worth, Mar. 2014, pp. 2237 - 2244.
[7] A. D. Napoli, F. Crescimbini, S. Rodo, and L. Solero, “Multiple input
dc-dc power converter for fuel-cell powered hybrid vehicles”, in Proc.

ICRERA 2014 513