IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO.
8, AUGUST 2013
Control Strategy for Power Flow Management in a
PV System Supplying DC Loads
Balasubramanian Indu Rani, Member, IEEE, Ganesan Saravana Ilango, and
Chilakapati Nagamani, Member, IEEE
Abstract—The growing concern for energy saving has increased
the usage of LED-based street lights, electronic chokes, compact
fluorescent lamps, and inverter-fed drives. Hence, the load profile
seen by the electrical grid is undergoing a notable change as
these devices have to operate from a dc source. Photovoltaics (PV)
being a major energy source, the aforementioned loads can be
connected directly to the dc bus. A grid-connected PV system
involves a power source (PV array), a power sink (load), and
two power sources/sink (utility and battery), and hence, a power
flow management system is required to balance the power flow
among these sources. One such system is developed for selecting
the operating mode of the bidirectional converter by sensing
the battery voltage. The viability of the scheme has been ascer-
tained by performing experimental studies on a laboratory proto-
type. The control strategy is digitally implemented on an Altera
Cyclone II Field Programmable Gate Array (FPGA) board, and
the algorithm is verified for different modes of operation by
varying the load. Experimental results are presented to bring out
the usefulness of the control strategy.
Index Terms—Bidirectional converter, dc bus, photovoltaic,
power flow management system (PMS).
I. I NTRODUCTION
W ITH THE ever-increasing demand for low-cost energy
and growing concern about environmental issues, PV-
based systems are being increasingly employed in diverse
applications both at domestic and commercial levels [1]. Pho-
tovoltaic systems can be broadly classified into stand-alone
system and grid-connected system [2], [3]. The stand-alone
system is widely used in remote places where access to elec-
tricity is not viable. The stand-alone configuration can provide
a well-regulated load voltage, but the reliability of power supply
cannot be guaranteed [3]. Storage batteries are widely used
to improve the reliability of the stand-alone system [4]. The
integration of a PV system to the grid is rapidly increasing due
to the improvement in the power electronics technology [5], [6].
Various topologies and control strategies [1]–[4], [7]–[28] for
grid-connected inverters have been reported in the literature. In
grid-connected PV systems (GCPVs) [13]–[24], the generated
PV power is fed to the grid, or it supplies the linear and
Manuscript received November 24, 2011; revised March 1, 2012 and
April 20, 2012; accepted May 25, 2012. Date of publication June 8, 2012; date
of current version April 11, 2013. This work was supported by the National
Mission on Power Electronics Technology (an initiative by Department of
Information Technology, Government of India) toward laboratory infrastructure
development.
The authors are with the Department of Electrical and Electronics Engineer-
ing, National Institute of Technology, Tiruchirappalli 620015, India (e-mail:
indu_b04@yahoo.com; gsilango@nitt.edu; cnmani@nitt.edu).
Digital Object Identifier 10.1109/TIE.2012.2203772
0278-0046/$31.00 © 2012 IEEE
3185
nonlinear loads connected at the ac side. A fair amount of
literature [3], [4], [25]–[29] has dealt with the operation of
hybrid systems. In some hybrid systems [4], [23] battery is
used to compensate the mismatch between the generation and
demand. In developing countries, the demand for power is
growing beyond the planner’s estimation, and the power grid
is weak, and scheduled power outages throughout the year
are common. In addition, there are also unscheduled short-
term outages which are random and frequent. As a result, a
grid-connected PV system with a battery backup has many
advantages such as peak shaving to generate power during peak
load hours, and therefore, the grid-side inverter should operate
in grid-tied mode and off-grid mode to supply uninterrupted
power to the critical loads during power outages [9]. In some
applications [15], [26], [29], a battery was connected directly
in parallel with the dc bus. The size of the battery can be
reduced when a battery charger/discharger circuit is inserted
between the dc bus and the battery [25], [30]. Utilizing the
battery charger/discharger circuit for regulating dc link voltage
decouples the dc link control from the ac current control and
achieves faster regulation of dc link voltage.
In the literature, much attention has not been paid to PV
systems supplying dc loads. The growing emphasis on energy
saving has increased the usage of LED-based street lights,
electronic chokes, compact fluorescent lamps, and inverter-fed
drives. PV being a major energy source enables the dc loads to
be connected directly to the dc bus. In this paper, the control
of a PV system feeding dc loads is considered for study. In
GCPVs, the variation in PV power will cause fluctuation in dc
link voltage, which is regulated by adjusting the ac line current
amplitude [3], [12]–[18], and under grid failure, a separate
controller is required to regulate the dc bus voltage [3], [25]. For
systems with dc loads, in addition to the PV power variation, the
fluctuation in loads also affects the dc bus voltage. The variation
in the dc link voltage causes a proportional change in the ac line
current. This causes the line current amplitude to vary often,
leading to poor power quality [22].
Moreover, when the solar radiation is less, the power quality
is degraded as the total harmonic distortion (THD) of the line
current is inversely proportional to the power injected into the
grid [25], [31]. Furthermore, the disturbance in dc link voltage
is reflected in PV array power, and the PV operating point
shifts from the maximum power point (MPP). Hence, a fast
regulation of dc link voltage is necessary to ensure constant
dc voltage while simultaneously maintaining the THD of the
current injected into the grid within the standard limits. An
attempt was made in [16] to provide reliable power to the loads
3186 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013

connected at the dc side, yet in this case, the PV power is not

utilized effectively. When the generated PV power is more than
the load demand, only the load power demand is met by the
PV, and the excess PV power is not fed to the grid. Moreover,
the impact of dc loads on the performance of the system is not
discussed.
The proposed PV system consists of three power sources
(grid, PV array, and battery), three power sinks (grid, battery,
and load), and a power flow management system (PMS) to
balance the power flow among these. This paper illustrates an
effective power flow management scheme for providing unin- Fig. 1. Grid-connected PV system with ac and dc loads.
terrupted power supply to the dc loads from a grid-connected
PV system while facilitating the evacuation of excess PV power
at high quality into the grid. Combining (6) and (7)
2T 2T T
2
Vdc 2
= Vdc(ref) − (pbo − pL ) + pb + Vm I m .
II. S YSTEM D ESCRIPTION Cdc Cdc Cdc η
(8)
In a grid-connected PV system with dc loads as shown in
From (8), it is clear that the fluctuations in PV power due to the
Fig. 1, the instantaneous power relationship is given by
change in solar radiation and the load power cause variations in
pbo (t) = pb (t) + pdc (t) + pL (t) + pinv (t) (1) the dc link voltage. The average current at the dc link can be
written as
where pbo is the output power of the boost converter, pb is the
power delivered by the charger/discharger circuit, pdc is the Ibo = IL + Iinv + Ib + Idc (9)
power to the dc link capacitor, pL is the power consumed by
where Ibo is the output current of the boost converter, IL is
the dc load, and pinv is the power extracted by the inverter. The
the load current, and Iinv is the current on the input side of
instantaneous ac power can be written as
the inverter. If Vpv and Ipv are the PV voltage and current,
pac = (Vm sin ωt)(Im sin ωt) (2) respectively, for the boost converter operating at a duty cycle D,
the following equations can be written:
Vm I m Vm I m
= − cos 2ωt (3)
2 2 (1 − D)Ipv = IL + Iinv + Ib + Idc (10)
where Vm is the amplitude of the phase voltage and Im is the Vpv
Vdc =   (11)
amplitude of grid current. The ac power includes a dc term and (1 − D) 1 + RL
R(1−D)2
a second-order ripple in the dc voltage. The average power on
the dc side can be written as where RL is the parasitic resistance of the inductor and R is the
equivalent resistance of the load. When there is a sudden change
Pinv = Vdc Iinv (4) in load, Vdc changes to (Vdc ± ΔVdc ). Since the perturbation
of MPP tracker (MPPT) is slow, D remaining the same, the PV
where Iinv is the average current on the dc side of the inverter.
voltage shifts from the optimal voltage
Equating the average power on the dc side to the dc term on the
ac side Vpv = Vmpp ± ΔVpv (12)
Vm I m Ipv = Impp ∓ ΔIpv (13)
= ηVdc Iinv (5)
2
Ppv = Vmpp Impp ± ΔVpv Impp ∓ ΔIpv Vmpp
where η is the efficiency of the inverter. If Vdc and Vdc(ref) are
the actual and reference values of dc link voltage, respectively, ∓ ΔIpv ΔVpv . (14)
the change in energy (ΔEdc ) stored in the dc link capacitor Cdc
It is clear from (11) and (14) that the disturbances in dc link
can be written as
voltage are reflected in the PV array power and shift the PV
Cdc  2 
power from the MPP until the dc voltage is restored. Hence,
ΔEdc = Vdc(ref) − Vdc2
. (6)
2 a fast regulation of dc link voltage is necessary to ensure a
constant dc link voltage. A charger/discharger controller is
To inject the PV power to the grid while maintaining a constant
designed based on (8) to regulate the dc link voltage. The
Vdc , the following energy balance should be satisfied:
control structure of the system is shown in Fig. 2 where Vdc
 
Vm I m is regulated by the charger/discharger circuit. The converter
ΔEdc = T pbo − pL − pb − (7) control to output transfer function Gdc [32] is expressed as (15),
2η
shown at the bottom of the next page, where rs is the resistance
where T is the time period of ac supply. of the battery, C1 and C2 are the input and output capacitances,
INDU RANI et al.: CONTROL STRATEGY FOR POWER FLOW MANAGEMENT IN PV SYSTEM SUPPLYING DC LOADS 3187

modes of operation is used to regulate the dc link voltage. The

PV system focuses on providing an uninterrupted power supply
to the loads connected at the dc side, and the grid is used
as a backup means when there is insufficient PV power and
complements the grid when excess PV power is available. This
necessitates the bidirectional converter to have three operating
modes, namely, idle mode (grid is disconnected), inverter mode,
and rectifier mode. The battery is charged either from PV or
grid depending upon the availability of the PV power and load
Fig. 2. Control structure of the proposed system. demand. The voltage control loop of the charger/discharger
circuit determines the charging/discharging mode of the battery.
TABLE I
PARAMETERS OF THE C ONTROLLER
The PV system involves three energy sources PV, battery, and
grid, and a PMS balances the power flow among these sources.
The control algorithm for the PMS is implemented on an Altera
Cyclone II FPGA board, and the control strategy is dealt in the
next section.

A. Control Strategy for PMS

respectively, L is the leakage inductance of the transformer, fsw
is the switching frequency, and d is the phase shift between the The PMS monitors the battery voltage and generates a
two bridges. The source to input transfer function [32] is given mode selection signal for the bidirectional converter, and it
by (16), shown at the bottom of the page. The transfer function also generates the reference current for the hysteresis current
of the load is represented by GL controller. The converter mode selection is carried out using
predetermined double bands of battery voltages. The outer
vdc (s) 1
GL (s) = =− (17) band corresponds to two voltage limits, i.e., the outer upper
IL (s) Cdc (s + 1/RCdc ) threshold voltage (VOUT ) and the outer lower threshold voltage
where R is the equivalent resistance of the load, and the relation (VOLT ). These limits correspond to the extreme voltages of
between PV current and the dc link voltage is expressed as the battery, i.e., the float value and the deep discharge level,
respectively. Similarly, the inner band voltage limits, i.e., the
vdc (s) inner upper threshold voltage (VIUT ) and inner lower threshold
Gpv (s) = = R(1 − D). (18)
Ipv (s) voltage (VILT ) correspond to interior voltage levels for the
purpose of control. For instance, for a battery with a nominal
Ginv denotes the transfer function of the grid-connected voltage of 36 V, the outer limits are VOUT = 40 V and VOLT =
inverter 32 V, and the inner limits are VIUT = 38 V and VILT = 34 V.
k vdc (s) Furthermore, the information of prevailing battery voltage (Vb )
Ginv (s) = = (19) and the present mode of converter are utilized in deciding
(1 + sτ ) Vg (s)
the next mode of the converter such that the converter does
where k is the amplitude gain and τ is the sampling delay. The not repeatedly shuttle between two different modes when the
parameters used in the controller are shown in Table I. system is operating near a threshold. The idea is to ensure that
The PV system consists of a PV array, a boost converter, the converter does not oscillate between rectifier and inverter
a battery storage unit with its charger/discharger circuit, and modes of operation.
a bidirectional converter to interface with the grid, and the The converter-mode-selection strategy is illustrated using
schematic of the PV system is shown in Fig. 3. A boost Table II and Fig. 4. The main feature of the strategy is that
converter is used to track the maximum power from the PV the transition from idle mode to conducting (rectifier/inverter)
array, and a dual H-bridge bidirectional converter (battery mode is based on the outer band limits, whereas a changeover
charger/discharger) capable of operating in both buck and boost from conducting mode to idle mode is based on the inner

vdc (s) (s + 2/rs C1 ) [(2d − 1)2V1 (3C2 − C1 )] /(fsw LC1 C2 )

Gdc (s) = = (15)
d(s) s + (2/rs C1 + 2/rs C1 )s + 4/Rrs C1 C2 + (2d2 − 2d)2 /fsw
2 2 L2 C C
1 2

vdc (s) (−2d2 + 2d)2/(f Lrs C1 C2 )

Gvb (s) = = (16)
Vb (s) s2 + (2/rs C1 + 2/rs C1 )s + 4/Rrs C1 C2 + (2d2 − 2d)2 /f 2 L2 C1 C2
3188 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013
Fig. 3. Schematic of the bidirectional converter with PMS.
TABLE II
C ONTROL S TRATEGY FOR PMS
band limits. The control circuit for the implementation of
PMS is shown in Fig. 5(a), and the variation of Vb and the
reference current under each case is shown in Fig. 5(b). Based
on the voltage threshold limits, two selection signals (S1 , S0 )
for the multiplexer determine the mode (M) of operation of
the converter. The inputs for the multiplexer 0, 1, and −1
denote the idle, rectifier, and inverter modes of operation,
respectively. The operation of the PMS is explained for the
conduction when Vb < VOLT . In this situation, the converter
has to switch to rectifier mode irrespective of the previous
Fig. 4. State diagram showing the power flow under various modes of
operation of the converter.
mode (idle or rectifier). However, the reference current for the
hysteresis controller is dictated by the previous mode. In case 1,
the reference current is the a predefined constant current Ic ,
whereas in case 8, it is (Ic + δI) (Table II). Next, if the battery
voltage increases such that VOLT ≤ Vb , the converter continues
to be in rectifier mode with the same reference current until
Vb ≤ VIUT and is indicated in case 7. Similarly, when the
battery voltage is higher than VOUT , then, the converter will
INDU RANI et al.: CONTROL STRATEGY FOR POWER FLOW MANAGEMENT IN PV SYSTEM SUPPLYING DC LOADS
Fig. 5. (a) Control circuit of the PMS. (b) Reference current under different
cases shown in Table II.
move to the inverter mode irrespective of the previous mode
(cases 3 and 5). However, the current reference is dictated by
the previous mode. Thus, it is ensured that the converter does
not oscillate between rectifier and inverter modes under any
situation.
Initially, assuming the converter to be in idle mode, the PV
supplies the load connected at the dc side and also charges
the battery when Ppv > PL (case 2). This increases the battery
voltage, and once Vb ≥ VOUT , the converter enters into inverter
mode. If Ppv < PL , then both battery and PV contribute to
supply the load. However, this decreases Vb , and when Vb ≤
VOLT , the converter enters into rectifier mode. The power flow
diagram for this mode is shown in Fig. 6(a).
When the converter is in inverter mode, the PV power is
injected into the grid, and the power flow diagram for this mode
is shown in Fig. 6(b). Here, there is a possibility of Vb to in-
crease or decrease depending on whether the battery is charging
or discharging. If VILT ≤ Vb ≤ VOUT , (case 4) the reference
current is injected into the grid. However, when Vb ≥ VOUT , the
reference current is incremented by a small amount δI (case 5)
such that Vb ≤ VOUT . Once Vb is less than the outer upper
threshold, there will be no change in the reference current, and
the converter operates in inverter mode until Vb > VILT after
which it stops injecting power and becomes idle (case 6).
When the converter is in rectifier mode, the battery may
charge or discharge depending on the availability of PV power
and load demand. If VOLT ≤ Vb ≤ VIUT , the current Ic is
3189
Fig. 6. Power flow under various modes of operation of the bidirectional
converter. (a) Idle mode. (b) Inverter mode. (c) Rectifier mode.
drawn from the grid. However, if the battery is discharged to
less than VOLT , the reference current is incremented by δI so
that Vb > VOLT . As seen from Table II (case 8), the change
in reference current is δI, and the PV and the grid charge the
battery to VIUT . At this condition, the converter enters into idle
mode (case 9), and the direction of power flow for this mode is
shown in Fig. 6(c).
B. Hysteresis Controller
A hysteresis current controller which is well proven and
adopted to many industrial applications [33] is employed for
the current control of an inverter. The reference signal for the
hysteresis current controller is generated as a function of the
phase of the grid voltage using Phase Locked Loop (PLL).
The injected current is in phase with the grid voltage, and
thus, unity power factor is achieved. For an average switching
frequency fsavg , the value of inductance [33] is calculated as
18 mH using

Vdc a2
L1 = 1− (20)
4Δifsavg 2
where Vdc is the dc link voltage, “a” is the modulation index,
and Δi is the ripple in the inverter current. The specifications
of the hysteresis current controller are shown in Table III.
3190 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013
TABLE III
S PECIFICATIONS OF THE H YSTERESIS C ONTROLLER
TABLE IV
S PECIFICATIONS OF THE E XPERIMENTAL S ETUP
III. R ESULTS AND D ISCUSSION
To evaluate the performance of the developed control strat-
egy, the time-domain simulation studies in MATLAB/Simulink
environment are presented, and the experimental tests are car-
ried out on a prototype unit built in the laboratory. The details of
the experimental setup are shown in Table IV. The system con-
sists of a PV array, a boost converter, a single-phase full-bridge
converter, and a charger/discharger circuit for the battery. The
array configuration includes five parallel strings with each hav-
ing two panels (rated 80 W · p) in series to get 36 V at the MPP.
A boost converter raises the PV voltage to 110 V. A bidirec-
tional dual H-bridge circuit which can operate in either buck or
boost mode acts as the charger/discharger circuit of the battery.
The well-known phase shift modulation strategy [34] is used to
control the charger/discharger circuit. The voltage control loop
of the charger/discharger circuit maintains the dc link voltage.
The single-phase bidirectional converter is built using power
Insulated Gate Bipolar Transistor ( IGBT) SKM100GB063D
and is connected to the grid through an 80/230-V line frequency
transformer. The gate pulses to the converter are fed through
SKHI22A driver circuits. The grid voltage and the battery volt-
age are sensed using isolated voltage transducers LV25P, and
the actual current is measured using current transducer LA55P.
The control algorithm including MPP tracking, dc link voltage
control, hysteresis current control, and the power management
system are digitally implemented in an Altera Cyclone II FPGA
board. The performance of the converter under steady state and
transient conditions such as step change in load and step change
in insolation is analyzed and compared with the performance of
the ac line current control technique [22]. The control strategy
of the PMS is verified for different modes of operation by
varying the load.
A. Steady-State Response
The steady-state response of the single-phase converter op-
erating in rectifier mode is shown in Fig. 7. It can be observed
from Fig. 7(a) that the grid voltage and line current are in phase,
Fig. 7. Steady-state response. (a) Rectifier mode. (b) Inverter mode.
indicating that the converter is operating in rectifier mode and
supplies 230 W at 0.99 power factor. The THD of the line
current is found to be 3.7%. Fig. 7(b) shows the steady-state
performance of the system in the inverter mode, and a 192-W
power is fed to the grid from the PV. The THD of the line
current in this mode is found to be 4.1%. It can be observed that
a near unity power factor is achieved in both the rectifier and
inverter modes of operation. In both the modes of operation, the
minimum current is set as 1 A by the PMS. Under any operating
mode, the injected current is more than the minimum value, and
therefore, the THD will always be less than the standard limit
of 5%.
B. Dynamic Response
To validate the effectiveness of the system in employing the
charger/discharger circuit to regulate the dc link voltage, the
dynamic performance of the system for two different cases,
namely, a step change in dc load and a step change in insolation,
is obtained and compared with the results obtained for a typical
ac line current control.
In a typical ac line current control, the mismatch between
input PV power and the load power dictates the mode of
operation of the converter, and the dc link voltage is regulated
by varying the ac line current magnitude. On the other hand, in
the proposed control, the battery voltage is used to determine
the mode of operation of the converter, and the dc link voltage
is regulated by the battery charger/discharger circuit. For a step
change in load, two different cases are considered.
INDU RANI et al.: CONTROL STRATEGY FOR POWER FLOW MANAGEMENT IN PV SYSTEM SUPPLYING DC LOADS 3191

Fig. 8. Dynamic response for step change in load: (a) AC line current control (iL —0.5 A/div; Vdc —55 V/div; Ppv —100 W/div; io —0.5 A/div). (b) AC
line current control (iL —1 A/div; Vdc —55 V/div; Ppv —100 W/div; io —0.5 A/div). (c) Proposed control (iL —1 A/div; Vdc —55 V/div; Ppv —100 W/div;
ib —1 A/div; io —1 A/div; Vb —5 V/div). Dynamic response for step change in insolation: (d) AC line current control (iL —0.5 A/div; Vdc —40 V/div; Ppv —100
W/div; io —1 A/div). (e) AC line current control (iL —0.5 A/div; Vdc —40 V/div; Ppv —100 W/div; io —1 A/div). (f) Proposed control (iL —0.5 A/div; Vdc —55
V/div; Ppv —200 W/div; ib —1 A/div; io —1 A/div; Vb —5 V/div). Time: 0.025 s/div.

1) When a step change in load closely matches the generated
PV power with the load power (Ppv ≈ PL ).
In this case with ac line current control, the remaining
less amount of power (Ppv − PL ) is fed to the grid. One
such situation is shown in Fig. 8(a). When the dc load
is increased from 164 to 202 W, with the PV power
remaining the same, the current injected into the grid is
reduced to 0.1 A. This increases the THD of the injected
line current to a high value of 28.8%. The magnified
view of ripple in the ac line current io is shown in the
subplot in Fig. 8(a). However, in the proposed technique,
the minimum current reference is set as 1 A by the PMS
as shown in Fig. 7(b).
2) When a step change in load changes (Ppv > PL ) to
(Ppv < PL ).
In this case, the converter initially operating in in-
verter mode changes to rectifier mode. Fig. 8(b) shows
that, initially, the inverter supplies the surplus PV power
(Ppv > PL ) to the grid, and as the load power is increased
to 342 W, the PV power (282 W) is insufficient to supply
the load demand, and the grid supplies the deficit power. It
is clear from Fig. 8(b) that fluctuations in load conditions
cause the converter to swing between inverter and rectifier
modes of operation. Moreover, it can be observed that
the disturbance in Vdc causes PV array power to decrease
from its maximum value.
In the proposed system, for the same step increase in
load from 224 to 342 W (Ppv < PL ), the increase in load
current is compensated by the increase in current ib of the
charger/discharger circuit. The current injected into the grid
remains constant as set by the PMS and is shown in Fig. 8(c). It
can also be observed that the change in battery voltage during
this period is less (39.2 to 37.7 V), and hence, the converter
does not change the mode of operation of the converter.
Similarly, for step change in insolation, two cases are
considered.
1) When a step change in insolation closely matches the
generated PV power with the load power (Ppv ≈ PL ).
In the ac line current control with the load power
remaining constant at 160 W, the PV power is decreased
to 348 W. The current injected into the grid is reduced to
0.7 A, and this increases the THD to 5.2%. The subplot
in Fig. 8(d) is the magnified view of the current io . It can
be observed that the THD of the current injected into the
grid varies with the insolation. In the proposed technique,
the charger/discharger complements the PV power, and
always, a minimum current is fed to the grid.
2) When a step change in insolation changes (Ppv > PL ) to
(Ppv < PL ).
In the ac line current control, the PV power is reduced
to 168 W (Ppv < PL ) with the load power remaining
3192
Fig. 9. Grid failure under (a) inverting mode and (b) rectifier mode. y-axis: Iinv : 4 A/div; Vb : 80 V/div; Ib : 4 A/div; IL : 4 A/div. x-axis: time: 50 s/div.
constant at 202 W. As shown in Fig. 8(e), the converter
changes from inverter to rectifier mode. With the pro-
posed system for the same step change in insolation, the
deficit power is supplied by the charger/discharger circuit,
and a minimum current set by the PMS is injected into
the grid as shown in Fig. 8(f). It can also be observed that
the change in battery voltage during this period is from
38.6 to 37.1 V, and hence, the converter does not shuttle
between different modes.
It can also be observed from the aforementioned results
that dc link voltage regulation using ac line current affects
the THD of the injected grid current, and the converter often
oscillates between the rectifier and inverter modes of operation
for variation in load and PV power. Moreover, the disturbance
in Vdc causes PV operation to shift from the MPP. Overall,
the proposed scheme is seen to be effective in preventing the
undesirable shuttling of the PV operating point and maintains
the THD of the injected grid current within the allowable limit
by setting a minimum current reference for injection.
C. Grid Failure Mode
When the grid power fails, irrespective of the present mode
of operation of the converter (rectifier/inverter), the PMS forces
the converter into idle mode. Both PV and battery supply the
load at the dc side. The worst case scenario under inverting
mode is grid failure when Vb is slightly less than VOUT and
Ppv > PL , and one such situation is shown in Fig. 9(a). With
233 W of PV power and 160 W of load power, the converter op-
erates in inverter mode and injects 230 W of power into the grid.
When Vb = 39.2 V, a sudden removal of grid supply forces the
converter to idle mode, and now, the power balance is to be
achieved among PV, load, and battery. As the excess PV power
charges the battery to VOUT , dump loads are switched on, and
both PV and battery together supply the load. This is evident
from the reversal of current direction in the charger/discharger
circuit.
The worst case scenario under rectifier mode occurs when
grid power fails when Vb is slightly more than VOLT and
Ppv < PL , and such a situation is shown in Fig. 9(b). With
228 W of PV power, 242 W of load power, and Vb = 32.6 V,
the converter operates in rectifier mode and charges the battery.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013
A sudden removal of grid supply forces the converter to idle
mode; as Ppv < PL , the battery supplies the deficit power;
when its voltage drops to 32 V, noncritical loads are removed to
avoid deep discharge of batteries; the other loads are supplied
by PV.
D. Automatic Power Flow Mode
An increase in PV power or a decrease in load power leads to
inverter mode, whereas a decrease in PV power or an increase
in load power leads to rectifier mode. Hence, to test the control
strategy of the PMS, four different cases are considered for
study as follows:
Case 1) inverter to rectifier mode via idle mode;
Case 2) inverter to inverter mode via idle mode;
Case 3) rectifier to inverter mode via idle mode;
Case 4) rectifier to rectifier mode via idle mode.
Since the mode of operation of the converter depends on
the battery voltage, to verify the control strategy, the variation
in battery voltage is obtained by changing the load across
the battery. The resistive loads are chosen such that a step
change in load causes the variation in battery voltage. The
experimental waveforms are shown in Fig. 10. Traces 1 and 2
show the battery voltage and battery current, respectively, and
the converter current at the dc side is shown in trace 3.
Case I and Case II: When the PV power (352 W) is more
than the load power requirement of 242 W, the converter is in
inverting mode and injecting 230 W of power into the grid.
The contribution of the battery charger/discharger circuit is
144 W. Now, an increase in dc load power (287 W) leads to
the reduction of battery voltage. This is obtained by applying
loads in steps at which the battery voltage falls below VILT
(33.8 V), and at this instant, the converter enters into idle mode,
indicating a zero current. An increase in load power (320 W)
leads to the reduction in battery voltage to 31.8 V, and the
converter enters into rectifier mode as shown in Fig. 10(a),
whereas a decrease in load power (222 W) makes the converter
to operate in inverter mode and inject 230 W of power to the
grid and is shown in Fig. 10(b).
Case III and Case IV: To verify Cases III and IV, prior to
the experiment, the loads are applied, and the battery voltage
is brought below 32 V, so that the power flow is from the grid
INDU RANI et al.: CONTROL STRATEGY FOR POWER FLOW MANAGEMENT IN PV SYSTEM SUPPLYING DC LOADS
Fig. 10. Battery voltage, battery current, and dc current waveforms for different cases under automatic power flow control. (a) Case I. (b) Case II. (c) Case III.
(d) Case IV. y-axis: Vb : 10 V/div; Ibat : 2 A/div; Iinv : 4 A/div. x-axis: time: 20 s/div.
to the dc side. With 297 W of PV power and 221 W of power
from the grid, the battery charger/discharger circuit charges the
battery. Now, the loads are released in steps, which rises the
battery voltage above the VIUT (38.3 V), and the converter
enters into idle mode. The removal of load increases Vb above
VOUT (40.2 V), and the converter changes into inverter mode,
and 230 W of power is fed to the grid. This can be observed
from Fig. 10(c). However, an increase in load demand reduces
Vb . This situation is verified by applying loads in steps and is
shown in Fig. 10(d). It is evident from the results that the system
exhibits a reasonably good response to the variation in load.
IV. C ONCLUSION
A versatile control strategy for power flow management in a
grid-connected PV system feeding dc loads has been presented.
The importance of the scheme has been brought out by per-
forming experimental studies on a laboratory prototype. The
steady-state performance of the converter for different modes
of operation has been observed, and near unity power factor
has been achieved in both the rectifier and inverter modes. The
transient performance of the system for step changes in load
and insolation have been also illustrated.
The control strategy has been digitally implemented on an
Altera Cyclone II FPGA board, and the algorithm has been ver-
ified for different modes of operation by varying the load, and
a good correlation between the results of computer simulation
and experiments has established the validity of the PMS.
3193
The significance of the proposed scheme has been demon-
strated by its effectiveness in preventing undesirable shuttling
of the PV operating point and also in maintaining the THD
of the injected grid current within the allowable limit of 5%
by setting a minimum current reference for injection. The
proposed configuration has been proved to be attractive from
the perspective of providing uninterruptible power to dc loads
while ensuring the evacuation of excess PV power of high
quality into the grid.
R EFERENCES
[1] Yazdani and P. P. Dash, “A control methodology and characterization of
dynamics for a photovoltaic system interfaced with a distribution net-
work,” IEEE Trans. Power Del., vol. 24, no. 3, pp. 1538–1551, Jul. 2009.
[2] X. Q. Guo and W. Y. Wu, “Improved current regulation of three-phase
grid-connected voltage-source inverters for distributed generation sys-
tems,” IET Renew. Power Gener., vol. 4, no. 2, pp. 101–115, Mar. 2010.
[3] H. C. Chiang, T. T. Ma, Y. H. Cheng, J. M. Chang, and W. N. Chang,
“Design and implementation of a hybrid regenerative power system com-
bining grid-tie and uninterruptible power supply functions,” IET Renew.
Power Gener., vol. 4, no. 1, pp. 85–99, Jan. 2010.
[4] F. Giraud and Z. M. Salameh, “Steady-state performance of a grid-
connected rooftop hybrid wind–photovoltaic power system with battery
storage,” IEEE Trans. Energy. Convers., vol. 16, no. 1, pp. 1–7, Mar. 2001.
[5] J. M. Carrasco, L. G. Franquelo, J. T. Bialasiewicz, E. Galvàn,
R. C. P. Guisado, M. A. M. Prats, J. I. León, and N. Moreno-Alfonso,
“Power electronic systems for the grid integration of renewable energy
sources: A survey,” IEEE Trans. Ind. Electron., vol. 53, no. 4, pp. 1002–
1016, Aug. 2006.
[6] B. Kroposki, C. Pink, R. DeBlasio, H. Thomas, M. Simões, and P. K. Sen,
“Benefits of power electronic interfaces for distributed energy sources,”
IEEE Trans. Ind. Electron., vol. 25, no. 3, pp. 901–908, Sep. 2010.
[7] H. M. Kojabadi, B. Yu, I. A. Gadoura, L. Chang, and M. Ghribi, “A
novel DSP-based current controlled PWM strategy for single phase grid
3194 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 60, NO. 8, AUGUST 2013
connected inverters,” IEEE Trans. Ind. Electron., vol. 21, no. 4, pp. 985–
993, Jul. 2006.
[8] J. C. Moreno, J. M. Espi Huerta, R. G. Gil, and S. A. Gonzalez, “A robust
predictive current control for three-phase grid-connected inverters,” IEEE
Trans. Ind. Electron., vol. 56, no. 6, pp. 1993–2004, Jun. 2009.
[9] Z. Yao, L. Xiao, and Y. Yan, “Seamless transfer of single phase grid
interactive inverters between grid connected and standalone modes,” IEEE
Trans. Power Electron., vol. 25, no. 6, pp. 1597–1603, Jun. 2010.
[10] M. Singh, V. Khadkikar, A. Chandra, and R. K. Varma, “Grid intercon-
nection of renewable energy sources at the distribution level with power-
quality improvement features,” IEEE Trans. Power Del., vol. 26, no. 1,
pp. 307–315, Jan. 2011.
[11] S. Yang, Q. Lei, F. Z. Peng, and Z. Qian, “A robust control scheme
for grid-connected voltage-source inverters,” IEEE Trans. Ind. Electron.,
vol. 58, no. 1, pp. 202–212, Jan. 2011.
[12] F.-S. Pai, J.-M. Lin, and S.-J. Huang, “Design of an inverter array for
distributed generations with flexible capacity operations,” IEEE Trans.
Ind. Electron., vol. 57, no. 12, pp. 3927–3934, Dec. 2010.
[13] H. Patel and V. Agarwal, “Investigations into the performance of
photovoltaics-based active filter configurations and their control schemes
under uniform and non-uniform radiation conditions,” IET Renew. Power
Gener., vol. 4, no. 1, pp. 12–22, Jan. 2010.
[14] Y.-C. Kuo, T.-J. Liang, and J.-F. Chen, “Novel maximum power point
tracking controller for photovoltaic energy conversion system,” IEEE
Trans. Ind. Electron., vol. 48, no. 3, pp. 594–601, Jun. 2001.
[15] Y.-K. Lo, T.-P. Lee, and K.-H. Wu, “Grid-connected photovoltaic system
with power factor correction,” IEEE Trans. Ind. Electron, vol. 55, no. 5,
pp. 2224–2227, May 2008.
[16] Y. K. Lo, H. J. Chiu, T. P. Lee, I. Purnama, and J. M. Wang, “Analysis
and design of a photovoltaic system connected to the utility with a power
factor corrector,” IEEE Trans. Ind. Electron., vol. 56, no. 11, pp. 4354–
4362, Nov. 2009.
[17] E. Romero-Cadaval, M. I. Milanés-Montero, E. González-Romera, and
F. Barrero-González, “Power injection system for grid-connected pho-
tovoltaic generation systems based on two collaborative voltage source
inverters,” IEEE Trans. Ind. Electron., vol. 56, no. 11, pp. 4389–4398,
Nov. 2009.
[18] J. He, Y. Li, and M. S. Munir, “A flexible harmonic control approach
through voltage-controlled DG-grid interfacing converters,” IEEE Trans.
Ind. Electron., vol. 59, no. 1, pp. 444–455, Jan. 2012.
[19] R. A. Mastromauro, M. Liserre, T. Kerekes, and A. Dell’ Aquila, “A
single-phase voltage-controlled grid-connected photovoltaic system with
power quality conditioner functionality,” IEEE Trans. Ind. Electron,
vol. 56, no. 11, pp. 4436–4444, Nov. 2009.
[20] J. M. Espí, J. Castelló, R. García-Gil, G. Garcerá, and E. Figueres, “An
adaptive robust predictive current control for three-phase grid-connected
inverters,” IEEE Trans. Ind. Electron., vol. 58, no. 8, pp. 3537–3546,
Aug. 2011.
[21] J.-M. Kwon, B.-H. Kwon, and K.-H. Nam, “Grid-connected photovoltaic
multistring PCS with PV current variation reduction control,” IEEE Trans.
Ind. Electron., vol. 56, no. 11, pp. 4381–4388, Nov. 2009.
[22] Y.-M. Chen, H.-C. Wu, Y.-C. Chen, K.-Y. Lee, and S.-S. Shyu, “The AC
line current regulation strategy for the grid connected PV system,” IEEE
Trans. Power Electron, vol. 25, no. 1, pp. 209–218, Jan. 2010.
[23] S. V. Araújo, P. Zacharias, and R. Mallwitz, “Highly efficient single phase
transformerless inverter for grid connected photovoltaic systems,” IEEE
Trans. Ind. Electron, vol. 57, no. 9, pp. 3118–3128, Sep. 2010.
[24] R. Kadri, J.-P. Gaubert, and G. Champenois, “An improved maximum
power point tracking for photovoltaic grid-connected inverter based on
voltage-oriented control,” IEEE Trans. Ind. Electron., vol. 58, no. 1,
pp. 66–75, Jan. 2011.
[25] S.-K. Kim, J.-H. Jeon, C.-H. Cho, J.-B. Ahn, and S.-H. Kwon, “Dynamic
modeling and control of a grid-connected hybrid generation system with
versatile power transfer,” IEEE Trans. Ind. Electron., vol. 55, no. 4,
pp. 1677–1688, Apr. 2008.
[26] E. Serban and H. Serban, “A control strategy for a distributed power
generation micro grid application with voltage and current controlled
source converter,” IEEE Trans. Power Electron, vol. 25, no. 12, pp. 2981–
2992, Dec. 2010.
[27] J. T. Bialasiewicz, “Renewable energy systems with photovoltaic power
generators: Operation and modeling,” IEEE Trans. Ind. Electron., vol. 55,
no. 7, pp. 2752–2758, Jul. 2008.
[28] T. Hirose and H. Matsuo, “Standalone hybrid wind–solar power gener-
ation system applying dump power control without dump load,” IEEE
Trans. Ind. Electron., vol. 59, no. 2, pp. 988–997, Feb. 2012.
[29] W. Qi, J. Liu, X. Chen, and P. D. Christofides, “Supervisory predictive
control of standalone wind/solar energy generation systems,” IEEE Trans.
Ind. Electron., vol. 19, no. 1, pp. 199–207, Jan. 2011.
[30] K. Jin, X. Ruan, M. Yang, and M. Xu, “A hybrid fuel cell power system,”
IEEE Trans. Ind. Electron., vol. 56, no. 4, pp. 1212–1222, Apr. 2009.
[31] M. Liserre, F. Blaabjerg, and S. Hansen, “Design and control of an LCL-
filter-based three-phase active rectifier,” IEEE Trans. Ind. Appl., vol. 41,
no. 5, pp. 1281–1291, Sep./Oct. 2005.
[32] H. Bai, C. Mi, C. Wang, and S. Gargies, “The dynamic model and hybrid
phase shift control of a dual active bridge converter,” in Proc. IECON,
2008, pp. 2840–2845.
[33] P. A. Dahono, “New hysteresis current controller for single-phase full-
bridge inverters,” IET Power Electron., vol. 2, no. 5, pp. 585–594,
Sep. 2009.
[34] C. Mi, H. Bai, C. Wang, and S. Gargies, “Operation, design and control of
dual H-bridge-based isolated bidirectional DC–DC converter,” IET Power
Electron., vol. 1, no. 4, pp. 507–517, Dec. 2008.
Balasubramanian Indu Rani (M’10) received the
B.E degree from Government College of Engineer-
ing, Tirunelveli, India, and the M.E degree from
Anna University, Chennai, India.
From 2004 to 2009, she was with the School of
Electrical Sciences, VIT University, Vellore, India.
She is currently a Full-time Research Scholar with
the National Institute of Technology, Tiruchirappalli,
India. Her areas of interest include renewable energy
systems and power electronics.
Ganesan Saravana Ilango received the B.E. degree
from the University of Madras, Chennai, India, in
2000, the M.E. degree from Bharathidasan Univer-
sity, Tiruchirappalli, India, in 2001, and the Ph.D.
degree from the National Institute of Technology,
Tiruchirappalli.
From 2001 to 2004, he was a Lecturer with Noorul
Islam College of Engineering, Kumaracoil, India.
Since 2006, he has been an Assistant Professor with
the National Institute of Technology. His preferred
working areas are Flexible AC Transmission Sys-
tems controllers, digital controllers, renewable energy systems, and power
quality.
Chilakapati Nagamani (M’10) received the B.Tech.
degree from the Sri Venkateswara University College
of Engineering, Tiruapati, India, the M.Tech. degree
from the Indian Institute of Technology, Kanpur,
India, and the Ph.D. degree from the University of
Technology, Sydney, Australia.
From 1985 to 1991, she was with the Central
Power Research Institute, Bangalore, India. Sub-
sequently, she joined the Department of Electrical
and Electronics Engineering, National Institute of
Technology (then known as Regional Engineering
College), Tiruchirappalli, India, as a Lecturer, where she is currently a Pro-
fessor. Her areas of interest include power electronics and drives, renewable
energy systems, and FACTS controllers.