Modeling and Quantitative Design of A Controller For A Bidirectional Converter With High Voltage Conversion Ratio
Modeling and Quantitative Design of A Controller For A Bidirectional Converter With High Voltage Conversion Ratio
Modeling and Quantitative Design of A Controller For A Bidirectional Converter With High Voltage Conversion Ratio
Abstract. The purpose of this paper is to perform the quantitative design of high voltage
conversion ratio bidirectional converter controller applied in photovoltaic power gener-
ation system. In order to keep the voltage dynamic response of load step change with
the features of zero steady-state error, no overshoot, small voltage-drop and fast restore
time, we perform the quantitative design of a controller by the given performance specifi-
cation. First, we use the step response estimation method at the selected operation point
to estimate the dynamic model of the converter. Then, we follow the given load regula-
tion performance to make the quantitative design of controller’s parameters, enable the
converter’s output DC link voltage be controlled at the planned dynamic and steady-state
response performance with stable energy supply. Finally, we use the measured results to
verify the performance of the designed controller.
Keywords: Photovoltaic power generation system, High voltage conversion ratio bidi-
rectional converter, Quantitative design, Step response estimation method, Load regula-
tion
2203
2204 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
to decide the proportional-integral (PI) controller parameters [2-6], which will make sys-
tem voltage unstable, or even unable to get the planned DC bus voltage. The voltage
controller designed by using Bode plots of frequency response [7,8] can find the voltage
controller parameters at the specific operation point, yet if the operation point shifts, it
is unable to get a better load regulation performance. At present we already have the
intelligent controller such as the fuzzy controller [9,10], sliding-mode controller [11-13] and
model predictive control [14], which can get better control performance, yet the controller
design is complicated, and with the stability and hard-to-implement issues. Some robust
and optimal controllers are presented in [15-18], but they still have the problem of com-
plex computation and implementation. Some of the controllers mentioned above are not
robust, and even if they are robust, they are not easy to implement because of the com-
plexity of computing. Therefore, to get better robustness and easy for implementation,
the quantitative performance design method is proposed for the proposed high voltage
conversion ratio bidirectional converter controller in this paper.
Based on this, in this paper, we use the step response estimation method to estimate the
dynamic model of converter. Then, follow the given load regulation performance specifi-
cations to perform the quantitative design of controller parameters, which controls the DC
link voltage of converter output at the planned dynamic and steady-state performance
without any unstable issue. In dealing with the current-controlled PWM scheme, the
small-signal mode of current loop is first found using the state-space averaging method,
and then, the current controller parameters are found based on frequency response. After
deriving the dynamic model of the proposed converter, a systematic design technique has
been adopted to find the parameters of a PI voltage controller according to the prescribed
specifications. The effectiveness of the proposed controller will be confirmed by simulation
and experimental results. Therefore, this controller can be systematized for quantitative
design and meets the required performance specifications.
In this paper, Section 2 describes briefly the operating theory of the proposed bidirec-
tional high voltage converter ratio converter. Then, the modeling and quantitative design
of a controller for the proposed converter are made in Section 3. In Section 4, some simu-
lation and experimental results are made to demonstrate the effectiveness of the proposed
controller. Finally, in Section 5, some conclusions are given.
MODELING AND QUANTITATIVE DESIGN OF A CONTROLLER 2205
Since the circuit of high voltage conversion ratio bidirectional converter has two-way-
flow ability, following the flow direction the energy can be divided into the boost and buck
operation modes; the circuit operations can be divided into the closed and open circuit
according to the switch actions. In an operating period, the ratio of closed time ton , the
ratio of opened time toff and period T within an operating period is defined as the duty
ratio D, expressed in Equation (1).
∆ ton ton
D= = (1)
T ton + toff
In the following section, we will illustrate the operation theory of high voltage conversion
ratio bidirectional converter by the boost mode and buck mode.
2.1. Boost mode.
(1) Switch-closed (0 5 ton 5 D1 T )
When the switch S1 at low-voltage side is closed, since the switch-control signals at
both the high voltage (HV) and low voltage (LV) side apply the complementary control
method, S2 at HV side is at off state; the responded equivalent circuit of the converter is
shown in Figure 3; now, the voltages measured at both ends of inductors L1 and L2 can
be expressed as
diL1
vL1 = VBat = L1 (2)
dt
n2
vL2 = VBat (3)
n1
where VBat is the voltage at low voltage side; n1 and n2 are turns of the coupled inductor
in primary and secondary sides, respectively.
If the turns ratio of coupled inductor N is defined as
∆ n2
N= (4)
n1
2206 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
Substituting Equation (16) into (15), we can derive the relation equation of voltages at
LV and HV sides as
VBat (1 − D1 )
= (17)
VBus (1 + N D1 )
From Equations (9) and (17) we can get that increasing the turns ratio can elevate the
voltage conversion ratio; the relation between the turns ratio and voltage to be suffered
while switch is off is expressed in the equation below [1]
(VBus + N VBat )
vDS 1 = VBat − vL1 = (18)
(1 + N )
vDS 2 = vL2 + VBus = (VBus + N VBat ) (19)
Currents taken while all switches are on (closed) are expressed in Equations (20) and
(21), where IBus is the DC link current, and IB is the battery’s current.
IDS 1 = IB + N IBus (20)
IB + N IBus
IDS 2 = (21)
(1 + N )
To have the high voltage conversion ratio bidirectional converter operated at the contin-
uous conduction mode (CCM), the design inductance must satisfy Equation (22) expressed
below [1]
VBat RBus (VBus − VBat )(1 − D1 )T
L1 min = 2
(22)
2VBus (N + 1)
where RBus is the load resistance at the high-voltage side of the DC voltage bus as shown
in Figure 1.
signal, it generates the pulse width modulation (PWM) signal, used to control the switch
state of power switches S1 and S2 . If voltage deviation occurs at HV side, tuning the value
and direction of battery current command i∗o over the voltage control loop can ensure that
the voltage at HV side is stabilized at the set command value.
3.1. Design of current controller. The current control of high voltage conversion ratio
bidirectional converter is to have the high voltage conversion ratio bidirectional converter
work at the continuous conduction mode (CCM) and make design [20,21]; in this paper,
we let the converter work at boost mode and make analysis on it; treat both L1 and L2
as an equivalent inductor L. Since the controlled battery current is equal to the current
at inductor L, while S1 at LV is closed, its state equation can be expressed as
diB (1 + N )vBat
=
dt L
dvCH vBus 0 ≤ t ≤ D1 T (23)
= − ,
dt R C
Bus H
vBus = vCH , iS2 = 0
After expressing Equation (23) in matrix, it becomes
[ ][
[ ] 0 0 ] (1 + N )
d iB iB
dt vC = 1 + L vBat
H
0 − v CH
RBus CH 0 (24)
[ ]
[ ] iB
vBus = 0 1 , iS2 = 0
vCH
While S1 at LV is opened, its state equation can be expressed as
diB vBat − vBus
=
dt L
dvCH iB vBus D1 T ≤ t ≤ T (25)
= − ,
dt CH RBus CH
vBus = vCH , iS2 = iB
After expressing Equation (25) in matrix, it becomes
[ ] 1 [ ]
0 − 1
d iB L iB
dt vC = 1 1 v + L vBat
H
− C H
0 (26)
[
CH
]
RBus CH
[ ] iB
vBus = 0 1 , iS2 = iB
vCH
Next, we use the state average method [22-25] to average the state equation in S1
switch’s duty cycle, i.e., Equation (24) × D1 T + Equation (26) × (1 − D1 )T and divided
by T , we obtain
(1 − D )
[ ] 0 −
1
[ ] 1 + N d1
d iB iB
=
L L vBat
dt vC (1 − D 1 ) 1 vCH +
−
H
0 (27)
CH RBus CH
[ ]
[ ] iB
vBus = 0 1 , iS2 = (1 − d1 )iB
v
CH
2210 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
To linearize Equation (27), we add the disturbing signal at the operation point, i.e., iB =
IB + ∆iB , vCH = VCH + ∆vCH , vBat = VBat + ∆vBat , vBus = VBus + ∆vBus , iS2 = IS2 + ∆iS2
and d1 = D1 + ∆d1 ; after putting the disturbing signal into Equation (27), we get
[1 − (D1 + ∆d1 )]
[ ] 0 − [ ]
d I + ∆i L I + ∆i
B B
=
[1 − (D1 + ∆d1 )]
B
VCH + ∆vCH
B
dt VCH + ∆vCH 1
−
CH RBus CH
1 + N (D1 + ∆d1 )
+ L VBat + ∆vBat (28)
0
[ ]
[ ]
IB + ∆iB
VBus + ∆vCH = 0 1
VCH + ∆vCH
IS2 + ∆iS2 = [1 − (D1 + ∆d1 )](IB + ∆iB )
Set the disturbing value of Equation (28) to zero, we can get that the steady-state op-
eration point is Equation (29), the disturbing signal equation at operation points (VBat ,
VBus , IS2 , IL , D1 ) is shown in Equation (30).
(1 − D1 )
VBat = VBus
1 + N D1
VBus = VCH = (1 − D1 )IB RBus (29)
V
IS2 = Bus , IS2 = (1 − D1 )IB
RBus
d∆iB (1 − D1 ) 1 + N D1 VC + N VBat
=− ∆vCH + ∆vBat + H ∆d1
dt L L L
d∆vCH (1 − D1 ) 1 IB (30)
= ∆iB − ∆v C − ∆d1
dt CH R C
Bus H
H
CH
∆V
Bus = ∆VCH , ∆iS2 = (1 − D1 )∆iB − IB ∆d1
After making Laplace transformation on Equation (30), we can plot the small signal
dynamic model block diagram, shown in Figure 8. By doing so, it can control the loop
current, where Ki is the current sensing factor.
B (s)
After deriving, the transfer function ∆i
∆d1 (s)
in Figure 8 can be the transfer
∆vBat (s)=0
function of Equation (31).
∆iB (s) (VBus + N VBat )RBus CH s + 2VBus + N VBat
= (31)
∆d1 (s) ∆vBat (s)=0 RBus CH Ls2 + Ls + (1 − D1 )2 RBus
The transfer function of current loop gain is
Gci (s) V̂1 Ki [(VBus + N VBat )RBus CH s + 2VBus + N VBat ]
tri
LG(s) = (32)
RBus CH Ls2 + Ls + (1 − D1 )2 RBus
where V̂tri is the amplitude of triangular carrier wave. The current controller Gci (s) applies
the PI controller shown in Equation (33).
KP i s + KIi
Gci (s) = (33)
s
MODELING AND QUANTITATIVE DESIGN OF A CONTROLLER 2211
Figure 8. Control system block diagram of the input current of high volt-
age conversion ratio bidirectional converter
The derivation of dynamic model and controller quantitative design are pretty difficult,
and we thus apply a simple general rule in the design process: the crossover frequency fc
is less than half of the switching frequency fs [26], as expressed in Equation (34).
1
fc < fs (34)
2
Since the integration gain has less influence to the crossover frequency, we thus select
KIi = 10000, and use the condition of Equation (33) to find out the range of KP i . Next,
we use Matlab software to analyze the Bode plots for the different loop gain frequency
responses corresponding to different KP i when KIi = 10000, shown in Figure 9. After the
analysis, we get the condition of KP i as
|LG(s = j2πfc )| KIi = 10000 = 1 (35)
fc = 25kHz
Here we set KP i = 10, and get the parameter of current feedback controller as
KP i s + KIi 10s + 10000
Gci (s) = = (36)
s s
3.2. Dynamic model estimation. Next, we use the step-response method [27] to es-
timate the dynamic model of converter; the system dynamic response of high voltage
conversion ratio bidirectional converter is expressed by the voltage loop control block
diagram shown in Figure 10, where Kv = 0.005 is the voltage sensing factor due to a at-
tenuation ratio 200:1 of the voltage sensor, Kpv is the conversion coefficient of the voltage
change due to a power disturbance at HV side, Gcv (s) is the voltage controller, and Gp (s)
is the transfer function of converter.
While estimating the dynamic mode by the step response estimation method, the design
steps are as follows.
2212 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
a) To estimate the dynamic mode estimation, the voltage controller was set as the pro-
portional control Gcv (s) = Kp = 3 at the operation point (VBus = 324V, P = 1000W)
to let the system operate in a closed loop control.
∗
b) Given a step voltage command (∆vBus = 0.2, i.e., the voltage vBus at HV side changes
324V → 364V), the changing status of measured vBus is shown in Figure 11; the steady-
state voltage is 360V.
c) At the same operating condition, grant an irradiance-change to make the change of
module array output power become ∆Ppv = 150W, i.e., Ppv changes 1000W → 850W,
the response of measured vBus is shown in Figure 12; the steady-state voltage is 318V.
Figure 10. Voltage loop control block diagram for high voltage conversion
ratio bidirectional converter
MODELING AND QUANTITATIVE DESIGN OF A CONTROLLER 2213
d) From voltage control loop block diagram shown in Figure 10, we can derive the transfer
∗
functions of Equation (37) for ∆vBus versus ∆vBus and Equation (38) for ∆vBus versus
∆Ppv .
∆vBus bGcv Kv ∆ c
Hv (s) = ∗
= = 1 (37)
∆vBus ∆p=0 s + a + bGcv Kv s + d
∆vBus −bKpv Kv ∆ c
Hp (s) = = =− 2 (38)
∆Ppv ∗
∆vBus =0 s + a + bGcv Kv s+d
where
Gcv = Kp (39)
2214 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
u2 > 0) of equations f1 (u1 , u2 ) and f2 (u1 , u2 ); then, we use Equation (49) to solve the
parameters KP and KI of DC link voltage controller:
(u1 + u2 ) − a u1 u 2
KP = KI = (50)
b b
4. Simulation and Experimental Results. Table 1 lists the electrical specifications
and component ratings of the implemented high voltage conversion ratio bidirectional
converter. This paper mainly focuses on the design of controller of the proposed converter.
As for the design procedure of the converter’s circuit element specification, the previous
paper has already presented in [2]; therefore, in view of the limited length of the paper,
it does not repeat here. The dynamic model of this converter is derived as shown in
Equation (51); we make test on the controller we have designed at the specified operating
point (output voltage vBus = 324V and load power Ppv = 1000W).
b 48.6
Gp (s) = = (51)
s+a s + 20.8255
Parameter Specification
Low voltage side voltage (VBat ) VBat = 50V
High voltage side voltage (VBus ) VBus = 380V
Switching frequency (f ) f = 25kHz
Power rating (Po ) Po = 1000W
Coupled inductance (L1 , L2 ) L1 = 20µH, L2 = 80µH
Turn ratio of coupled inductor (N ) N =2
Power switches (S1 , S2 ) IXGH48N60C3D1 (600V/48A)
Capacitance (CL , CH ) CL = CH = 470µF
the traditional PI controller, although the voltage-dip has insignificant difference from the
one shown in Figure 14, the restoring time takes longer apparently.
While the photovoltaic module array power output is down for 250W, the measured
dynamic response of output voltage is shown in Figure 16. Although the 250W power step
change is in excess of the designed one, which makes the voltage-dip a little bit higher,
it is still better than the control performance of traditional PI controller shown in Figure
17, in particular on the brilliance of restoring time.
MODELING AND QUANTITATIVE DESIGN OF A CONTROLLER 2217
5. Conclusions. In this paper we build a high voltage conversion ratio bidirectional con-
verter, which can sustain a stable DC link voltage output when the photovoltaic module
array output power varies, and supplies sufficient power to load while the irradiance is
insufficient; while the irradiance is sufficient, it can store the extra power into battery.
Meanwhile, the voltage controller applying quantitative design enables DC link voltage to
have high regulation performance. The high voltage-ratio bidirectional converter made in
2218 P.-Y. CHEN, K.-H. CHAO AND H.-J. CHEN
this paper also can be applied in different renewable energies; the applications are wide
with practical value. A non-isolated soft-switching bidirectional converter will be studied
in the future to provide not only a high conversion ratio, but also a low voltage stress
across power switches. Furthermore, the surge voltage can be suppressed through the
release of the energy stored in leakage inductance. In addition, a robust PI voltage con-
troller with a quantitative design procedure will be presented to overcome a wide range
of load changes.
REFERENCES
[1] Y. M. Chen, H. C. Wu and Y. C. Chen, DC bus regulation strategy for grid-connected PV power
generation system, Proc. of IEEE International Conference on Sustainable Energy Technologies (IC-
SET), pp.437-442, 2008.
[2] B. L. Narasimharaju, S. P. Dudey and S. P. Singh, Coupled inductor bidirectional DC-DC converter
for improved performance, Proc. of IEEE International Conference on Industrial Electronics Control
and Robotics (IECR), pp.28-33, 2010.
[3] B. L. Narasimharaju, S. P. Dudey and S. P. Singh, A novel voltage clamped CI-BDC converter
and its control for battery storage applications, Proc. of IEEE International Conference on Power
Electronics, Drives and Energy Systems (PEDES), 2010.
[4] P. Tyagi, V. C. Kotak and V. P. S. Singh, Design high gain DC-DC boost converter with coupling
inductor and simulation in PSIM, International Journal of Research in Engineering and Technology
(IJRET), vol.3, pp.156-163, 2014.
[5] J. C. Lopez, M. Ortega and F. Jurado, New topology for DC-DC bidirectional converter for hybrid
systems in renewable energy, International Journal of Electronics, vol.102, pp.418-432, 2015.
[6] J. C. S. Júniora, J. P. R. Balestero and F. L. Tofoli, Novel bidirectional DC-DC converters based on
the three-state switching cell, International Journal of Electronics, vol.103, pp.878-897, 2015.
[7] B. L. Narasimharaju, S. P. Dudey and S. P. Singh, Voltage mode control of coupled inductor bidi-
rectional DC to DC converter, Proc. of IEEE International Conference of Electron Devices and
Solid-State Circuits (EDSSC), pp.1-6, 2010.
MODELING AND QUANTITATIVE DESIGN OF A CONTROLLER 2219
[8] B. L. Narasimharaju, S. P. Dudey and S. P. Singh, Parasitic effect considerations in modeling of CI-
BDC converter and its voltage controller, Proc. of International Conference on Power Electronics,
Drives and Energy Systems (PEDES), 2010.
[9] A. U. S. Jyothi, D. S. P. Gopal and G. Ramu, Bi-directional DC-DC converter drive with PI and
fuzzy logic controller, International Journal of Advanced Research in Electrical, Electronics and
Instrumentation Engineering (IJAREEIE), vol.2, pp.5435-5441, 2013.
[10] N. F. N. Ismail, I. Musirin, R. Baharom and D. Johari, Fuzzy logic controller on DC/DC boost
converter, Proc. of IEEE International Conference on Power and Energy (PECON), pp.661-666,
2010.
[11] A. Romero, L. Martinez-Salamero, H. Valderrama, O. Pallas and E. Alarcon, General purpose sliding-
mode controller for bidirectional switching converters, Proc. of IEEE International Symposium on
Circuits and Systems, pp.466-469, 1998.
[12] D. S. Oh, Y. J. Kim, J. D. La and Y. S. Kim, The sliding mode controller of the bidirectional
converter for the DC bus stabilization, Proc. of International Conference on Electrical Machines
and Systems (ICEMS), pp.367-371, 2010.
[13] A. T. Laura, V. I. Enric, M. A. Javier, M. P. Sylvia and M. S. Luis, Seamless sliding-mode control
for bidirectional boost converter with output filter for electric vehicles applications, IET Power
Electronics, vol.8, pp.1808-1816, 2015.
[14] P. Akter, M. Uddin, S. Mekhilef, N. M. L. Tan and H. Akagi, Model predictive control of bidirectional
isolated DC-DC converter for energy conversion system, International Journal of Electronics, vol.102,
pp.1407-1427, 2015.
[15] M. Takahashi, Self-repairing PI/PID control against sensor failures, International Journal of Inno-
vative Computing, Information and Control, vol.12, no.1, pp.193-202, 2016.
[16] S. Gao, X. Wu, G. Wang and G. Zhang, The immune fuzzy PID stripper temperature control algo-
rithm based on chaotic optimization, International Journal of Innovative Computing, Information
and Control, vol.12, no.1, pp.103-112, 2016.
[17] F. Liu and H. Wang, Fuzzy PID tracking controller for two-axis airborne optoelectronic stabilized
platform, International Journal of Innovative Computing, Information and Control, vol.13, no.4,
pp.1307-1322, 2017.
[18] R. Kurokawa, N. Inoue, T. Sato, O. Arrieta, R. Vilanova and Y. Konishi, Simple optimal PID tuning
method based on assigned robust stability -Trade-off design based on servo/regulation performance-,
International Journal of Innovative Computing, Information and Control, vol.13, no.6, pp.1953-1963,
2017.
[19] D. Buvana and R. Jayashree, Current control in two level non-isolated bidirectional DC-DC converter
in CCM for HEV, Proc. of IET Chennai 4th International Conference on Sustainable Energy and
Intelligent Systems (SEISCON), pp.228-239, 2013.
[20] D. W. Hart, Introduction to Power Electronics, Pearson Education, New York, 1997.
[21] F. H. Zhang, C. H. Zhu and Y. G. Yan, The controlled model of bi-directional DC-DC converter,
Proc. of the Chinese Society for Electrical Engineering (CSEE), vol.25, pp.46-79, 2005.
[22] R. D. Middlebrook and S. Cuk, A general unified approach to modeling switching converter power
stages, Proc. of IEEE Power Electronics Specialists Conference, pp.18-34, 1976.
[23] R. D. Middlebrook, S. Cuk and W. Behen, A new battery charger/discharger converter, Proc. of
IEEE Power Electronics Specialists Conference, pp.251-255, 1978.
[24] S. P. Hsu, A. Brown, L. Rensink and R. D. Middlebrook, Modelling and analysis of switching DC-
to-DC converters in constant-frequency current-programmed mode, Proc. of IEEE, 1979.
[25] C. K. Tse and K. M. Adams, Quasi-linear modeling and control of DC-DC converters, IEEE Trans.
Power Electronics, vol.7, pp.315-323, 2002.
[26] R. Paul, L. Sankey, L. Corradini, Z. Popovic and D. Maksimovic, Power management of wideband
code division multiple access RF power amplifiers with antenna mismatch, IEEE Trans. Power
Electronics, vol.25, pp.981-991, 2009.
[27] K. H. Ang, G. Chong and Y. Li, PID control system analysis, design, and technology, IEEE Trans.
Control Systems Technology, vol.13, pp.559-576, 2005.
[28] C. M. Liaw, Design of a two-degree-of-freedom controller for motor drives, IEEE Trans. Automatic
Control, vol.37, pp.1215-1220, 1992.