1
Response Coordination of Distributed
Generation and Tap Changers for Voltage
Support
An D.T. Le, Student Member, IEEE, K.M. Muttaqi, Senior Member, IEEE, M. Negnevitsky, Member,
IEEE,and G. Ledwich, Senior Member, IEEE
Abstract-- The recent introduction of the competitive
electricity market in many countries has sparked a renew trend in
connecting small-size generators into distribution networks.
Those new generators together with different other types of
equipment such as On-load Tap Changing (OLTC) transformer,
shunt capacitors, shunt reactors, etc, will all participate into the
voltage regulation process in the power systems. Poor
coordination between these devices may cause unnecessary
operations, and consequently unnecessary wear, unnecessary
energy consumption as well as poor voltage quality. In this paper,
we present an innovative strategy to coordinate the voltage
control actions in a distribution system with more than one
voltage regulating device. The method for voltage control
coordination is developed based on the priority level of each
regulating device and implemented through communication. A
sensitivity-based technique for determining the control zones of
the regulating devices has been developed. A practical system
with tap changers and distributed generator has been chosen to
test the developed control method. Simulations have been carried
out extensively on a practical distribution system to show the
effectiveness of the method.
Index Terms-- Distributed Generation, Power Distribution
System, Voltage Control, Control Zone, On-load Tap Changing
Transformer.
I. INTRODUCTION
O Ne of the most important duties of a distribution system is
to maintain the voltage supplied to customers within the
acceptable limits, under any loading circumstances. Failure to
meet this responsibility may result in malfunctions or damage
of customers’ equipment. Voltage fluctuations occurring in the
network is essentially the result of changing loads due to
varying demands of customers at different times throughout
the day. In order to keep these fluctuations within the statutory
limits, indirect method with distributed generations (DG),
shunt reactors/capacitor banks, Static VArs Compensators
This research has been financially supported by the Australian Research
Council under ARC Linkage Grant K0014223 “Integration of Distributed and
Renewable Power Generation into Electricity Gird Systems”, collaboration
with Aurora Energy, Tasmania.
Ms. An D.T. Le, Dr. Mohammad A. Kashem, and Prof. Michael
Negnevitsky are with the School of Engineering, University of Tasmania,
Hobart, Australia, and Prof. Gerard Ledwich is with the School of
Engineering Systems, Queensland University of Technology, Brisbane,
Australia, (emails: M.Kashem@utas.edu.au; dtale@utas.edu.au;
Michael.Negnevitsky@utas.edu.au; g.ledwich@qut.edu.au).
(SVC), etc., and direct method with On-load Tap Changing
(OLTC) transformer, regulators, etc. are used [1].
In this paper, the voltage regulation task is accomplished by
using the OLTC in conjunction with the DG. While the OLTC
is already available in most distribution feeders, the DG is
chosen here as voltage regulatory equipment for the many
benefits such as voltage support, line loss reduction,
improvement of supply reliability and security. A problem that
arises in this kind of voltage control system, however, is the
possibility of excessive operations of the OLTC and/ or the
DG. This might happen due to insufficient voltage control
method or a lack of proper coordination between the two
devices in their simultaneous operation. Unnecessary actions
of the OLTC (or voltage regulator) and DG are undesirable
because of economic reasons. Changing tap position of the
OLTC causes transients and mechanical wear on the OLTCs
themselves [2]. Operation of the DG consumes energy
resources (some of them are expensive, such as fuel) and also
reduces the operational age of the machine.
The voltage control coordination is therefore necessary in
the distribution network and has been a subject of interest in
many research papers. H.M. Ma et al in [3] have used the
hierarchical genetic algorithm (HGA) to optimise the power
voltage control systems according to the number of control
actions. In [4], an integrated voltage control called
Coordinated Secondary Voltage Control (CSVC) has been
proposed for controlling the OLTC positions to ensure that
voltage and loading constraints are satisfied during normal and
emergency conditions. Another voltage regulation method in
power distribution systems including DG systems has been
developed in [5] through optimizing sending voltages using the
Least Square method. Authors in [6] have coordinated the
operations of switched capacitors and OLTC in a radial
distribution system by approximating the problem as a
constrained discrete quadratic optimization. In [7], the
coordination method for the DG and step voltage regulator
operations for improved distribution system voltage
regulations has been presented. However, none of these
coordination methods has considered the priority/selectivity of
different voltage regulating devices in the system with the
presence of DG. This concept is incorporated in this paper for
the purpose of increasing the effectiveness as well as reducing
the operating cost of the control actions.
The aim of this paper is to build up a complete, sufficient
and applicable voltage regulation method for the OLTC and

the DG in their simultaneous operation during peak load hours.
The objective of the coordination controller is to reduce, or
eliminate the number of counter-acting operations between the
two devices.
II. VOLTAGE REGULATION METHOD
Conventionally, the voltage on any distribution feeder is
controlled by the OLTC placed at the substation. The main
purpose of this voltage regulation is to keep the voltages of all
the customers within the acceptable limits, which is ± 5% for
Australia distribution system. However, with the rapidly
increase of load, the OLTC can not always guarantee that it
will be able to keep the regulation requirement satisfied. The
recent trend of introducing DG into the distribution networks
has become a potential solution for solving voltage problem,
especially during peak hour. In this section, we introduce a
coordinated voltage control technique using parallel operation
of the OLTC and the DG with minimum level of
communication system.
A. Simple System Model
To analyze the operation of the OLTC and the DG, as well
their effects on the distribution feeder, a simplified model has
been used as shown in Fig.1.
DG
~ The counter basically starts accumulating the time when
OLTC Load
d three cases: the voltage error fluctuates below and above the
L deadband, the voltage error changes sign, or the voltage error
Fig. 1 Simplified model of distribution feeder
The distribution feeder is connected to the substation,
which is modelled as a Thevenin source, at the point of
common coupling. For simplicity, distribution line of length L
is represented as series resistance and reactance with load
represented by equivalent lumped values at N number of
nodes. The network is equipped with an OLTC at the
substation. The OLTC has provision for correcting the voltage
from -5% to +15%, each step represents a 1.25% change in
voltage. A DG connected to the distribution line through a
load bus d km from the substation. It is modelled as a current
source and able to inject both real and reactive power to the
system.
B. Voltage Regulation by OLTC
The operation of the OLTC includes five major steps, which
are described in details below:
Step 1 - Status determination: The OLTC recognizes the
current voltage condition, which is represented by the value of
voltage error. The voltage error of the OLTC controller, Verr1
depends on two inputs, the actual voltage at the target bus of
the OLTC, Vt1, and the reference voltage Vref. It can be
calculated using (1):
2
Verr1 = V ref − Vt1 (1)
It should be noted that Vref is normally set at 1.0 p.u. The
target bus of the OLTC is referred to as the furthest bus from
the substation which belongs to the control zone of the OLTC.
In other words, the OLTC is meant to regulate the weakest
voltage point within their zone of responsibility. Definition of
“control zone” will be discussed later on in this paper. Step 1
produces an output called Si, which takes a value of 0, or -1, or
+1 depending on how difference the Verr1 compared to the
deadband a is, as shown in (2).
 0 for - a ≤ Verr1 ≤ a

S i = + 1 for Verr1 > a (2)
 - 1 for V < −a
 err1
Value of the deadband defines the total voltage range,
around the voltage setting Vref, which the OLTC controller will
consider as a satisfied condition. Choice of a small deadband
will result in more tap changes to occur, but will provide a
better regulation. On the contrary, a larger deadband causes
fewer tap changes, but at the expense of a closely regulated
line.
Step 2 – Comparison of current and previous voltage status: in
this step, the current status Si will be compared to the previous
status Si-1. This is done to keep track of the changes of voltage
status at the target bus.
Step 3 – Counter setting: A counter is employed to determine
the time duration of the voltage error exceeding the deadband.
voltage error is outside of the deadband. It will be reset in
is currently within the deadband. Output of the counter is
determined based on the results of the comparison in Step 2:
 0 for S i ≠ S i −1

Ci =  0 for S i = 0 (3)
C i −1 + ∆t for S i = S i −1
Where ∆t is the time step.
Step 4 – Time delay setting: In this step, value of the counter
will be compared to the time delay setting DT, and the control
action will take place only if the time delay is elapsed. The
time delay varies depend on the control coordination
algorithm, which has a great impact on the performance of the
system. Selection of the DT value will be discussed in more
details in the section of “Coordinated voltage control”.
Step 5 – Action of tap changer: if there is a requirement of a
tap change and the counter value is equal or greater than the
time delay setting (Ci ≥ DT), a signal will be sent to the tap-
changer motor to move the tap up or down. The actual tap
ratio is:
ni = ni −1 + u (∆n )
(4)
if n min ≤ n i ≤ n max and V min ≤ Vlocal ≤ Vmax
Where ni and ni-1 are the current and previous actual tap ratio,
respectively, ∆n is the tap step, and u is the desired tap ratio
which is +1 for requirement of a tap up, -1 for requirement of a
tap down, and 0 for no requirement of tap action. nmin and nmax

represent the working band of the OLTC, Vmin and Vmax are the
lower and upper voltage limits, respectively.
C. Voltage Regulation by DG
The DG controller operates in a much similar manner with a
sequence of 5 steps. The steps and their outputs are as
followings:
Step 1 - Status determination: The DG status SDGi can be
determined by (5);
 0 for - b ≤ Verr 2 ≤ b
 Where,
S DGi = + 1 for Verr 2 > b (5)
 - 1 for V
 err 2 < −b
and Verr 2 = V ref − Vt 2 (6)
Where, Verr2 is voltage error of the DG controller, Vt2 is actual
voltage at the target bus of the DG, b is the voltage deadband
of Vt2.
Step 2 – Comparison of current and previous voltage status:
SDGi is compared to SDGi-1.
Step 3 – Counter setting: Counter value is calculated as:
 0 for S DGi ≠ S DGi −1
 s DG = −Y3−1Y5
C DGi = 0 for S DGi = 0 (7)
C DGi −1 + ∆t for S DGi = S DGi −1
Step 4 – Time delay setting: Time delay of the DG is denoted
as TDDG. Setting for time delay of the DG controller will be
described later in section 2.4.
Step 5 – Action of the DG: The desired output of the DG, IDG,
is calculated using the proportional controller as in (8):
I DG = K P V err 2 (8)
Where KP is the proportional constant.
In this work, however, the DG is set to step-by-step update
its output towards to desired value, instead of making a
substantial change. This is done to allow time for the other
regulating devices (OLTC in this case) to take part into the
regulating process. Also, it helps to avoid unnecessary extra
energy delivered by the DG in case when the later change of
the load demand would possibly eliminate the voltage
violation. By controlling DG output in this way, we would be
able to reduce the operation cost of the DG. However, it
causes increasing time of the DG response to voltage violation
if the current load condition maintains sufficiently long.
D. Coordinated Voltage Control
In order to avoid hunting between the units, proper
coordination is needed. The coordination makes sure that all
regulating devices in the system will response to any voltage
violation in a correct order. As we have mentioned earlier, the
task of coordinating responses of the OLTC and DG in this
paper is done using a definition of “control zone”.
Idea of the control zone is that the area subject to voltage
control is divided into two zones. Each device is mainly
responsible for controlling voltage in one zone, where its
control action has higher influence to the voltage. The other
device will act as a supporter in case the action of the main is
not satisfactory or fails.
3
In this work, we propose a method for determination of
voltage control zones according to the sensitivity analysis.
From the network equation YBUSVBUS = IBUS, for a N+1-load
bus system with bus 1, bus n, and bus n+1 represent the
substation, the remote load bus, and the internal bus which
connects the DG into the system, respectively, we obtain:
(
V3− N = −Y3−1 Y2TVS − 2 + Y5VDG ) (9)
V3 
 y1,3 K y1, n 
=  M  , Y2 = 
y 2,3 K y 2, n 
V3 − N

V n 
V 
V S − 2 =  S  , and VDG is voltage at the DG connection point.
V 2 
The relative changes of V3-N due to the DG injection and the
tap operation can be obtained with substituting VS = 0 into (9)
and examining individually the response of the system to 1.0
p.u. voltage at the DG connection point and 1.0 p.u. voltage at
V2, which is the secondary side of the OLTC, using the
superposition principle. The results are:
(10)
0
s TX = −Y3−1Y2T   (11)
1
If values of (10) & (11) are graphically plotted with respect
to the load bus from 3 to N, their intersection point could be
defined as the boundary for control zones of the OLTC and the
DG. Let the ordinate of the intersection is Cb. The
responsibility of each device is defined by following rules:
- If the voltage sensitivity of bus i with respect to DG voltage
is greater than Cb ( s DG ≥ C b ), bus i is belong to the control
i
zone of DG.
- If the voltage sensitivity of bus i with respect to the V2 is
i
greater than Cb ( s tx ≥ C b ), bus i is belong to the control zone
of the OLTC.
The main purpose of voltage control zone determination is
that we will be able to assign higher opportunity to a more
efficient device to operate. Thus, the voltage drop will be
compensated more quickly with less number of control actions.
Let us assume that the boundary bus is bus m. In other
words, the control zone of the OLTC is from bus 2 to bus m,
while DG control zone is from bus (m+1) to bus n, which is
the remote bus. Delay time values of the OLTC and DG, which
are denoted as DT and DTDG, respectively, are not constant,
but changing from time to time depending on the network
conditions. The delay time could be one of the following
types:
- Transient time (TT for OLTC and TTDG for DG): the time
that has to elapse before any voltage correction is permitted to
take place. This setting is to avoid voltage correction to occur
due to temporary voltage excursions.
- Grading time (GT for OLTC and GTDG for DG): the time in
which a regulating equipment has to delay its action to enable
 4

the equipment with higher priority to complete their voltage if VD post − VD pre ≤ w
correction. In this paper, since the operation cost of the OLTC
is less than that of the DG, we assume that the default Grading if TD ≤ TT then TD = TT
Time of the DG is two times higher than the OLTC’s Grading else then TD = ST
Time. end
- Second-action time (ST for OLTC and STDG for DG): the else
time delay before any further action can take place, if required. if the DG fails to operate
The next action has to be in the same direction with the if TD ≤ TT then TD = TT
previous one for the Second-action time to apply. Otherwise, else then TD = ST
the time delay with be either Transient Time or Grading Time.
end
It should be noted that Second-action Time is always shorter
else
than Transient Time.
The algorithm for the control coordination has been if TD ≤ TT DG + TT then TD = TTDG + TT
proposed with a set up of communication system as follows: else then TD = ST
- One way communication from the target load buses to its end
corresponding controller, which is depending on the control end
zones. end
- Two way communication betweens controllers of regulating end (12)
devices in the system.
The first part of the communication system is built to provide Time delay for the DG:
highly accurate feedback signal to the controller, which is the
if VD post − VD pre ≤ w
voltage status at target buses. The later part of the
communication system, on the other hand, is used to inform if the OLTC fails to operate
status of one device to another. This would be very helpful in if TD DG ≤ TT DG then TD DG = TT DG
the case that when one device fails to operate, their
responsibility could then be passed on to other equipment else then TD DG = ST DG
within the shortest time, and thus reduce the duration of end
voltage rise/drop. end
The priority of controller in this coordination scheme has else
been determined based on the idea of control duty. This means if TD DG ≤ TTDG then TD DG = TTDG
that if more significant load change occurs in the control zone
of one regulating device, it then has more priority to operate else if TTDG < TDDG ≤ GTDG then TDDG = GTDG
than the other. However, taking into account that operation of else then TD DG = STDG
the DG is more expensive than the OLTC, the DG, in case of
end
higher priority, will only issue one step change in its output
end (13)
then wait until the Grading Time has passed before doing any
Where, VDpre and VDpost are the simulated voltage drops
further voltage correction. This will allow time for the OLTC
before and after voltage excursion. ω is a threshold value to
to operate if necessary. Moreover, by using the communication
determine the priority of the regulating equipment.
system, a signal could be sent from one controller to another to
In (12) & (13), the three variables VDpre, VDpost, and ω are
inform their status. If one device is exceeding its capacity or
used to determine the knowledge of where the causes of
fails to carry out the desired action, the other one will operate
voltage violation are. This knowledge is used as a criterion to
with minimum delay time, regardless its priority.
assign the time delay values for the OLTC as well as the DG.
The time delay of the OLTC and the DG for this control
When the remote voltage is outside the deadband, this is due to
coordination is calculated as in (12) & (13) using “IF”
one or both of the following reasons:
statements. It should be noted that the equations applied only if
- A voltage change upstream the DG
Si ≠ 0 and SDGi ≠ 0, respectively. Vr and Vm are used to denote
- A load change downstream the DG
the actual remote voltage and the actual voltage at boundary
The distinction between those two can be found by monitoring
bus m, respectively.
the local voltage and current at DG point, as well as the
Time delay for the OLTC:
upstream feeder impedance. This information could provide us
if Si ≠ 0
a simulated voltage drop, from which we are able to define if
if TD ≤ TT then TD = TT there is any load change has occurred in the voltage control
else then TD = ST zone of the DG [8]. If the difference between simulated
end voltage drop before and after voltage excursion is large, then
end the second cause is more likely to occur.
if SDGi ≠ 0
 5

III. CASE STUDY Determination of control zone

1
The proposed coordination method has been tested on a s DG
0.9
system, which is modified from a distribution feeder of Aurora s TX
Energy. The main feeder is 48 km long with 1.0136 MVA load 0.8

distributed along the line. Loads are assumed to be lumped at 0.7

Sensitivity values
69 nodes, in which node 1 is represented for the substation and
0.6
node 69 is denoted for the remote load bus. There are nine
laterals attached to the main, which is shown in Fig.2. 0.5

WOOLNORTH Control zone of the OLTC Control zone of the DG

0.4

0.3
H F 0.2
E
0.1

D C 0
G B 12 22 32 42 52 62 72 82 92
Load bus number
A Fig. 3: Determination of the control zones
SMITHTON
The load variation, as shown in Fig.4, has been used to test
Fig. 2: Smithton - Woolnorth test feeder the proposed method of voltage control coordination. Total
In order to properly illustrate how the coordination methods simulation time is 400 seconds. During this period of time, the
work, modified form of the system has been used. The feeder total load demand increases from 1.02 MVA up to 4.02 MVA.
has an automatic OLTC at the substation and a DG 36 km At each second, there are a number of random load buses
from the substation. changing their load demand with a random magnitude. This
has been done to imitate the practical situation of load
IV. SIMULATION RESULTS fluctuations, which are random and non-uniform. Since under-
voltage situations are more likely to occur in the distribution
Firstly, we demonstrate how to determine the control zones
systems, an increasing load profile has been used in the
for each regulating unit by the approach of zone determination
simulation. By doing this test, we would be able to know how
with the system offline load data. The sensitivity values of load
much the maximum loading capacity is before the system
bus voltages with respect to voltage at the secondary side of
voltage fails to meet the safety margins.
the OLTC (sTX) and with respect to voltage of the DG
connection point (sDG) have been plotted on the same graph, as Real and reactive power of load demand
4.5
shown in Fig.3. The intersection of sTX and sDG divides the
feeder into two parts. The upstream part from bus 2 to bus 52 4
Real power
is belong to the OLTC control zone, while the downstream Reactive power
3.5
part from bus 53 to the remote end belongs to the DG control
P (MW) and Q(MVars)

3
zone. We should note that the sensitivity analysis here is
carried out for those load buses located on the main feeder 2.5
only. For the laterals, those attach to the backbone at points
2
which belongs to the OLTC control zone, will also be the
major regulating duty of the OLTC. Otherwise, the DG will be 1.5
the main responsible regulating unit.
1

0.5

0
0 50 100 150 200 250 300 350 400
Time (sec)
Fig. 4: Dynamic load profile

Normally, the time delay of an automatic OLTC would be

around 20 secs for the first tap and 5 secs for the consequent
ones. However, since our simulation time is relatively short,
we assume that the time delay values are as follows:
TT = 10 secs, GT = 10 secs, and TT = 2 secs
TTDG = 3 secs, GTDG = 20 secs, and TTDG = 1 sec
In all simulations, the reference voltage Vref is 1.0 p.u. The
variables a and b are set to be 0.025p.u. and 0.045 p.u., which
 6

represent the variation limits of bus 52 and bus 69,
respectively. Since bus 52 is at around two-third of the feeder
from the substation, we would like to keep its voltage at a
reasonable level (±0.025 variation is chosen here) in order that
further voltage drops along the line will not cause the remote
end to fall below 0.95 p.u. The variable b is chosen to be 0.045
instead of 0.05 to give some levels of tolerance to the control
system.
Figs.5, 6 and 7 show the results of the coordination method
using communication system, including the tap response, the
DG response, and the voltage at the remote end, respectively.
In Fig.5, we can see that the OLTC alters its ratio according to
the voltage variation at bus 52. At first, it increases with the
increase of load. After that, at t = 220 seconds, tap ratio stops
rising at the value of 1.063. Even though the capacity of the
OLTC has not exceeded yet, further increase of tap ratio is not
allowed because of upper voltage constraint at the secondary
side of the OLTC.
not sufficient enough to activate the DG controller. When t =
352 seconds, the DG starts responding to the excursion at the
remote bus. After that, from t = 360 seconds, the DG output
increases rapidly to compensate the voltage drop as it detects
the saturation condition of the OLTC.
Voltage profile of system shows that the voltage is well
regulated most of the simulation time. Actions of the OLTC
has maintained the voltage within the safety margins, and thus
kept the DG offline for more than 85% of time. The DG only
participates into the control process when the voltage drops
below the limit, and the OLTC fails to operate further. This
can be considered as a desired solution for solving voltage
problems in a distribution system, which maximises the
capacity of the OLTC and utilises the DG only in real need.
Voltage at the remote end
0.99
0.985
0.98

Tap position 0.975

1.08
Voltage (p.u.)

0.97
Tap ratio / Voltage at target bus (p.u.)

1.06
0.965

1.04
0.96

1.02 0.955
Tap ratio
Actual voltage at bus 52 0.95
1
0.945
0 50 100 150 200 250 300 350 400
0.98 Time (sec)
Fig. 7: Remote end voltage profile
0.96

V. CONCLUSION
0.94
0 50 100 150 200 250 300 350 400 In this paper, an algorithm for the control action
Time (sec)
coordination between the OLTC and the DG has been
Fig. 5: Tap response and voltage at the target bus – bus 52
developed with the utilization of communication system. The
Real and reactive power of DG
0.18 method works by incorporating several concepts, including the
control zones, settings of reference voltage, voltage deadband,
0.16
as well as time delay. Selection of these settings has been
0.14 discussed in details in the paper, which makes it easier for the
control engineers to design their systems. The coordination
P (MW) and Q(MVars)

0.12 Real power

Reactive power method offers a reasonable recovery time for the system
0.1 voltage. At the same time, it tries to minimize the operating
0.08
cost of regulating devices by maximizing the capacity of the
OLTC, which is cheaper, and minimizing the control action of
0.06
the DG. Simulations have been conducted and the results
0.04 prove that the algorithm works well and also very flexible. The
approach can be expanded to be applicable for more complex
0.02
systems with more number of regulating devices, such as
0
0 50 100 150 200 250 300 350 400
voltage regulator. With the rapid growth of new technologies
Time (sec) and the decrease in their cost, it is believed that in the near
Fig. 6: DG response future, communication system will become popular and
affordable for most of utilities. As the result, the proposed
By observing the voltage at the remote end (Fig.7), which is
method can be used widely for satisfactory operation of the
controlled by the DG, we can see that the voltage drops below
distribution networks.
the deadband at t = 335 seconds. However, the DG has not
responded at that time since the duration of voltage violation is
 7

VI. ACKNOWLEDGMENT Member of CIGRE AP36 (Electromagnetic Compatibility), Australian

Technical Committee.
The authors gratefully acknowledge the support and
cooperation of Aurora Energy personnel in providing data and Gerard Ledwich (M’73–SM’92) received the Ph.D. in electrical
engineering from the University of Newcastle, Australia, in 1976. He has
advice on the operation of distribution systems.
been Chair Professor in Electrical Asset Management at Queensland
University of Technology, Australia, since 1998. He was Head of electrical
VII. REFERENCES engineering at the University of Newcastle from 1997 to 1998. Previously, he
was associated with the University of Queensland from 1976 to 1994. His
[1] M. Larsson, “Coordination of Cascaded Tap Changers using a Fuzzy-
interests are in the areas of power systems, power electronics, and controls.
Rule Based Controller”, Fuzzy Sets and Systems, Vol. 102, Issue 1,
Prof. Ledwich is a Fellow of the Institution of Engineers Australia.
pp.113 – 123.
[2] G.W. Kim, and K.Y. Lee, “Coordination Control of ULTC Transformer
and STATCOM Based on an Artificial Neural Network”, IEEE
Transactions on Power Systems, May 2005, Vol. 20, Issue 2, pp. 580 –
586.
[3] H.M. Ma, K.F. Man, and D.J. Hill, “Control Strategy for Multi-
objective Coordinate Voltage Control Using Hierarchical Genetic
Algorithms”, IEEE International Conference on Industrial Technology,
ICIT 2005, 14-17 Dec. 2005, pp. 158 – 163.
[4] F.A.B. Lemos, L.C. Werberich, J.S. Freitas, and M.A. da Rosa, “A
Strategic for Voltage Coordinated Control to Improve System
Operation”, 2001 IEEE Porto Power Tech Proceedings, 10-13 Sept.
2001, Vol. 2.
[5] D. Rho, H. Kita, J. Hasegawa, and K. Nishiya, “A Study on the Optimal
Voltage Regulation Methods in Power Distribution Systems
Interconnected with Dispersed Energy Storage and Generation
Systems”, Proceedings of EMPD '95., 1995 International Conference
on Energy Management and Power Delivery, 1995, 21-23 Nov. 1995,
Vol. 2, pp. 702 – 707.
[6] R. Baldick and F.F. Wu, “Efficient Integer Optimization Algorithms for
Optimal Coordination of Capacitors and Regulators”, IEEE
Transactions on Power Systems, Aug. 1990, Vol. 5, Issue 3, pp. 805 –
812.
[7] L.A. Kojovic, “Coordination of Distributed Generation and Step
Voltage Regulator Operations for Improved Distribution System
Voltage Regulation”, 2006 IEEE Power Engineering Society General
Meeting, 18-22 June 2006.
[8] C.A. Smith, M.A. Redfern, and S.Potts, “Improvement in the
Performance of On-Load Tap Changer Transformers Operating in
Series”, IEEE Power Engineering Society General Meeting, 13-17 July,
2003, Vol. 3.

VIII. BIOGRAPHIES
An D.T. Le received the B.E. (Hons.) from the University of Tasmania,
Australia, in 2004. She is currently pursuing the PhD degree at the University
of Tasmania. Her special fields of interests are power system analysis,
renewable energy, distributed generation, power system control and
protection.

K.M. Muttaqi (A’00–SM’05) received the Ph.D. degree from Multimedia

University, Malaysia, in 2001. Currently, he is a Senior Lecturer at the School
of Engineering, University of Tasmania, Australia. He was associated with
the Queensland University of Technology, Australia as a Postdoctoral
Research Fellow from 2000 to 2002. Previously, he also worked for
Multimedia University as a Lecturer for three years. His special fields of
interests include distributed generation, renewable energy, distribution system
automation, power system planning, and artificial intelligence. He has
published more than 50 technical papers in these areas.

Michael Negnevitsky (M’95) received the B.S.E.E. (Hons.) and Ph.D.

degrees from Byelorussian University of Technology, Minsk, Belarus, in
1978 and 1983, respectively. Currently, he is Professor, Chair in Power
Engineering and Computational Intelligence, in the School of Engineering at
the University of Tasmania, Hobart, Australia. From 1984 to 1991, he was a
Senior Research Fellow and Senior Lecturer in the Department of Electrical
Engineering, Byelorussian University of Technology. After arriving in
Australia, he was with Monash University, Melbourne, Australia. His
interests are power system analysis, power quality, and intelligent systems
applications in power systems. Dr. Negnevitsky is a Chartered Professional
Engineer, a Senior Member of the Institution of Engineers Australia, and a