Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

Visualization and Animation of State Estimation Performance

A. P. Sakis Meliopoulos, George J. Cokkinides Mike Ingram, Sandra Bell, Sherica Mathews
Georgia Institute of Technology Tennessee Valley Authority
[email protected]

Abstract the model utilized generates biases. These biases have

resulted in poor performance and several other practical
Reliable real time system “visibility” depends on a problems [7]. The experience is that the State Estimation
reliable and accurate state estimator. Present experience problem does not have 100% performance, i.e. there are
with state estimators indicates that its reliability is below cases and time periods that the SE algorithm will not
expectations (an average of 5% non-convergent cases). converge. The state estimator can be drastically improved
The causes of this poor performance have been identified with GPS synchronized measurements. Specifically,
in earlier work by the authors and alternative robust state recent technology of disturbance recorders introduced
estimators have been proposed. For any state estimator, it synchronized measurements. Synchronization is achieved
is important to develop techniques for monitoring the via a GPS (Global Positioning System) which provides
performance of the state estimator and identification of the synchronizing signal with accuracy of better than 1
potential problems such as bad sensors, consistent errors, µ sec . This time precision is translated into a precision
modeling errors, etc. The paper presents visualization and of 0.02 degrees of the US power frequency (60 Hz).
animation methods that assist this process. It is Therefore, the technology provides a means to measure
demonstrated that bad data many times can be detected the phase angles with a precision of 0.02 degrees. It
via visualization methods. The methodologies are basically provides a direct measurement of the system
demonstrated with a hybrid three-phase state estimator state. As such, it has been greeted by some as the
which addresses the issue of systematic errors from replacement of state estimators. This assertion however is
modeling and imbalance errors. This estimator is false. GPS synchronized measurements are imperfect
enhanced by visualization and animation methods that measurements as any other measurement. While the GPS
provide valuable information to users and system synchronized equipment may have higher precision than
operators “at a glance”. The procedure and the conventional metering, there are additional sources of
visualization techniques are demonstrated on TVA’s 500 error from the instrumentation channel, calibration errors
kV transmission system. and systematic errors introduced by the design of the
equipment (for example a constant shift of frequency). In
1. Introduction addition there are systematic errors from model
inaccuracies and random errors from failed data
A reliable and accurate real time model of a power system acquisition systems, transmission errors, etc. Filtering of
is of paramount important for effective control and all of these problems requires use of state estimators.
operation of the system. The real time visibility of the In an earlier paper we propose an approach that uses data
system is achieved with the use of the SCADA system from various sources (SCADA as well as GPS-
and processing of the SCADA data via state estimators. synchronized measurements) for the purpose of enhancing
The end result of this process is expected to be a reliable the state estimator. The methodology is based on a three
and accurate real time model of the system. The phase detailed power system model. The detailed three
importance of this process has become abundantly clear phase model is very important and contributes to the
during the August 14, 2003 blackout. Historically, the performance of the hybrid state estimator. This method is
importance of this issue was recognized immediately after utilized here in a variety of ways with emphasis on
the 1965 blackout. Following the 1965 blackout, power providing maximum information to system operators via
system state estimators were implemented in the late 60s visualization and animation methods. We focus on
to achieve this objective. The initial implementation was visualization methods that reveal bad data in the
based on single phase measurements and a power system measurement set, visualization of system imbalances,
model that is assumed to operate under single frequency, power system operating status and voltage profile issues.
balanced conditions and symmetric system model. These Numerical experiments are presented that illustrate the
assumptions are still prevalent today. The single benefits of the visualization methods using the hybrid
frequency, balanced and symmetric system assumptions state estimator. The numerical experiments are performed
have simplified the implementation but have generated a on the TVA 500 kV system.
disconnect between the model and the real time data, i.e.

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 1
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

2. Description of the Hybrid Three-Phase

2
State Estimator § z − hi ( x ) ·
Min J = ¦ ¨¨ i ¸¸
This section presents a short description of the hybrid i © σi ¹
three-phase state estimator. This state estimator uses a
combination of standard SCADA data and synchronized where σi is the meter accuracy.
data which are related to a full three phase system model
to perform state estimation.
Solution methods for above problem are well known. In
subsequent paragraphs, the models of the measurements
The state of the system is defined as the phasors of the
and the details of the hybrid state estimator are described.
phase voltages at each phase of a bus, including the
neutral node. A bus k may have three to five nodes,
phases A, B and C, possibly a neutral (N) and possibly a Table 1. List of Measurements
ground node (G). The state of the system at this bus is the
node voltage phasors. The system model is a three phase GPS-Synchronized Measurements
model with explicit representation of the neutral nodes Description Type Code
and ground nodes if present. For a four node bus, we use ~ 1
Voltage Phasor, V
the symbols A, B, and C for the phase nodes and N for the ~
Current Phasor, I 2
neutral node. The states are defined with:
~ 3
~ Current Injection Phasor, I inj
Vk , A = Vk , A,r + jVk , A,i
~
Vk , B = Vk , B ,r + jVk , B ,i Non-Synchronized Measurements
~ Description Type Code
Vk , A = Vk , A,r + jVk , A,i Voltage Magnitude, V 4
~
Vk , A = Vk , A,r + jVk , A,i Real Power Flow, Pf 5

Reactive Power Flow, Qf 6

In compact form, the state for a four node bus k will be:
Real Power Injection, Pinj 7
~
ªVk , A º ªVk , A,r + jVk , A,i º Reactive Power Injection, Q inj 8
«~ » «
~ «Vk , B » «Vk , B ,r + jVk , B ,i »
Vk = ~ = »
«Vk ,C » «Vk ,C ,r + jVk ,C ,i » 3. Description of the Measurement Data Set
«~ » « »
¬«Vk , N ¼» ¬Vk , N ,r + jVk , N ,i ¼
The available data in a power system can be classified
into (a) GPS synchronized measurements (phasors
The voltages of all buses of the system form the state of measurements) and (b) non-synchronized measurements.
the system. We will refer to this as the state of the system, A typical list of measurements has been given in Table 1.
x. Since at each bus the model may have a neutral node as
well as a ground node, the measured phase voltages are
The measurements can be GPS-synchronized always considered as the phase to neutral voltages. As it
measurements or usual SCADA data. A typical list of has been mentioned, the measurements are related to the
measurement data is given in Table 1. The measurements state of the system via the “model” equations. The state of
are assumed to have an error that is statistically described the system has been defined in the previous section.
with the meter accuracy. Each measurement is related to Figure 1 illustrates some typical measurements. The
the state of the system via a function. model equations, i.e. the equations that relate the system
state to the measurement are given below. The variables
Given a set of measurements, the state of the system is that appear in these equations are defined in Figure 1.
computed via the well known least square approach.
Specifically, let z i be a measurement and h i ( x ) be the
function that relates the quantity of the measurement to
the state of the system. The state is computed from the
solution of the following optimization problem.

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 2
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

Bus m ~
~ Vm
­ ~ *
device d2
Id1,m,A § ªV k , A º · ½
Pd1,m,A
Qd1,m,A ° ¨ « ~ »¸ °
° ¨ «V k , B » ¸ °
° ¨ «V~k ,C » ¸ °°
°~ ¨ T
A B C Qd1,k,A ced
1 Q d 1,k , A = Im ®V k , A C d 1,k , A « ~ » ¸ ¾
i ¨ «V m, A » ¸ °
Qd2,k,A ~ Pd1,k,A dev ° ¨ «V~m , B » ¸ °
Id1,k,A
Pd2,k,A °
~ ° ¨ « ~ »¸ °
Id2,k,A A B C ¨ «¬V k ,C »¼ ¸¹ °
°¯ © ¿
Bus k
~ A B C
Vk
It is also important to note that normally measurements of
neutral or ground voltages are not available. On the other
A B C
hand these voltages are very small under normal operating
conditions. For this reason, we introduce one pseudo-
measurement of voltage phasor for each neutral and
ground node in the system. The value of this measurement
is exactly zero. The “meter accuracy” for this
measurement is assumed to be low. Typically a value of
10% is used.
Figure 1. Measurement Definition – Three Phase
Model 4. Description of the Hybrid Three-Phase
~ ~ ~ State Estimator
zV ,k , A = Vk , A = Vk , A − V k , N
~ ~ ~ The hybrid three-phase state estimator uses standard
zV ,k , B = V k , B = Vk , B − Vk , N
SCADA data and synchronized data together with a full
~ ~ ~
zV ,k ,C = V k ,C = V k ,C − V k , N three phase system model to estimate the system state.
The measurement data has been discussed in the previous
~ section. The mathematical procedure is described next.
ªV k , A º
«~ » The measurements are assumed to have an error that is
«V k , B » statistically described with the meter accuracy. Thus, each
~ «V~k ,C » one of these measurements has the following
I d 1, k , A = C d 1,k , A « ~ » , similar equation for phases B
T
mathematical model.
«V m, A »
«V~m, B » Phasor measurements:
«~ »
«¬V k ,C »¼ ~ ~
zν = V k , A − V k , N + η~ν
~
and C.
~
~ ªVk , A º
Vk , A = V k , A = V k2, A, r + V k2, A,i «~ »
«Vk , B »
~ * ~ «V~k ,C » ~
­ § ªV k , A º · ½ ~
zν = I d 1,k , A + ην = C d 1,k , A « ~ » + ην
T
° ¨ « ~ »¸ ° «Vm , A »
° ¨ «V k , B » ¸ ° «V~m , B »
° ¨ «V~k ,C » ¸ °° «~ »
°~ ¨ T
Pd 1, k , A = Re ®V k , A C d 1,k , A « ~ » ¸ ¾ ¬«Vk ,C ¼»
° ¨ «V m, A » ¸ °
° ¨ «V~m , B » ¸ °
¨ Pseudo-measurements for neutrals and grounds:
° « ~ »¸ °
¨ «¬V k ,C »¼ ¸¹ ° ~
°¯ © ¿ zν = 0 + j 0 = V k , N + η~ν
~

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 3
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

non-synchronized measurements. The best state estimate

Non-synchronized measurements: is given by:

~ ~ 2 Case 1: Phasor measurements only.

zν = Vk , A − Vk , N + 2ην =
= (Vk , A,r − Vk , N ,r ) + (Vk , A,i − Vk , N ,i ) + 2ην xˆ = (H sT WH s ) H sT Wz s
2 2 −1

Case 2: Phasor and non-synchronized measurements.

­ § ~ *
½
ªVk , A º · xˆν +1 = xˆν +
° ¨ « ~ »¸ °
° ¨ «Vk , B » ¸ °
z s − H s xˆν
+ (H T WH ) H T W «
° ¨ ~ ¸ ° −1 ª º
zν = Pd 1,k , A
° ~ ¨ T «Vk ,C » ¸
+ ην = Re ®Vk , A C d 1,k , A « ~ »
¨
°
¾ + ην
ν ν T
ν »
¬ z n − H n xˆ − xˆ Qi xˆ ¼ { }
° «Vm , A » ¸ °
° ¨ «V~m, B » ¸ ° where:
° ¨ « ~ »¸ °
¨ «¬Vk ,C »¼ ¸¹
°¯ © °¿ ªW 0º ª Hs º
W =« s », H =« »
~ * ¬ 0 Wn ¼ ¬ H n + H qn ¼
­ § ªVk , A º · ½
° ¨ « ~ »¸ °
¨ 5. Visualization and Animation
° «Vk , B » ¸ °
° ¨ ~ ¸ °
° ~ ¨ T «Vk ,C » ¸ ° The data available from SCADA, GPS-synchronized
zν = Qd 1,k , A + ην = Im ®Vk , A C d 1,k , A « ~ » ¾ + ην
° ¨ «Vm , A » ¸ °
measurements and state estimation results are
¨ overwhelming to system operators in the usual tabular
° «V~m, B » ¸ ° reports of numerical values on a single line diagram.
° ¨ « ~ »¸ °
¨ ¸ Recent efforts resulted in displaying power flow data in 2-
°¯ © ¬«Vk ,C ¼» ¹ °¿ D or 3-D visualizations of the data [13], [14].
Visualization methods are powerful in enhancing the
comprehension of system operating conditions for users
The state estimation problem is formulated as follows: and system operators. It is important to use the same
technology for the results of state estimators for the
purpose of enhancing the information transfer. An
η~ν*η~ν ην ην effective visualization method will help users and system
Min J = ¦ + ¦
ν ∈ phasor σ ν ν ∈non − syn σ ν
2 2 operators to identify problems with one glance at the
displays. The types of information that is important are
(but not limited to this list):
It is noted that if all measurements are synchronized the
state estimation problem becomes linear and the solution
1. Are all measurements good? Are there any bad data?
is obtained directly. In the presence of the non-
2. Can I trust the computed real time model of the system?
synchronized measurements and in terms of above
3. What is the true operating condition of the system?
formulation, the problem is quadratic, consistent with the
quadratized power flow. Specifically, using the quadratic
The answer to these questions can be quickly assessed
formulation, the measurements can be separated into
with a number of displays. At the research level, we have
phasor and non-synchronized measurements with the
generated a matrix of useful information. The selection
following form:
matrix is shown in Figure 2. Note that the user may select
from a variety of quantities included the measurements,
zs = H s x +η s estimated values, residuals or normalized residuals. The
z n = H n x + {x T Qi x}+ η n
quantities may be voltage magnitudes at any phase (or
neutral), real or reactive power flow at any phase as well
as electric current magnitude or phase at any phase of the
In above equations, the subscript s indicates synchronized system. The user may select a single quantity or multiple
(phasor) measurements while the subscript n indicates

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 4
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

quantities and generate a visualization of this information 6. Numerical Experiments with the Hybrid
in a 3-D or 2-D display. State Estimator
The hybrid state estimator has been extensively tested
with numerical experiments. The numerical experiments
consist of a data generator that generates a set of
measurement data from a solved three phase power flow
condition. For the numerical experiments, a user selected
random error is added to the data. This section describes
the numerical experiments. The test system is the 500 kV
TVA system shown in Figure 4. The model includes the
entire TVA 500 kV system and the transformers and
autotransformers to the lower kV levels (mainly 161 kV
and some 230 kV). The remaining system (beyond the
secondary of the included transformers) is represented by
equivalents. The total number of nodes for this system is
1167 nodes.
Figure 2. Illustration of the Selection Matrix for
Visualizations Several scenarios have been studied that demonstrate the
effectiveness of the visualization methods for providing
An example of the developed visualization modules with useful information at a glance. Any problems are
3-D capability is illustrated in Figure 3. The visualization immediately identified. Here we present three simple
display shows the single line diagram of the system along scenarios. At the presentation of the paper these scenarios
with indicators of user selected quantities, such as as well as any other can be demonstrated live.
estimated or measured values of bus voltages (in the case
of Figure 3, magnitude (tubes) and phase (pies) residuals), Scenario 1: In this scenario it is assumed that the
circuit power and current flows, measurement errors following measurements are available: (a) real and
(computed as the difference between estimated and reactive power flow at the terminals of all circuits, all
measured values), and estimation residuals. All indicators phases, and (b) voltage phasors of each phase at all buses.
are analog so that extreme values (large errors etc) can be
easily spotted. For example, voltage magnitudes are Scenario 2: In this scenario it is assumed that the
indicated by vertical cylinders of height proportional to following measurements are available: (a) real and
the voltage. Phase angles are indicated by pie charts. reactive power flow at the terminals of all circuits, phase
Circuit flows are indicated by walls along the circuit lines A only, and (b) voltage phasors of phase A at all buses.
of height proportional to the flow, etc.
Scenario 3: In this scenario it is assumed that the
measurement data are identical to those of scenario 1
except that a large error has been added to one datum (50
MW in one flow measurement).

The measurement data for the above scenarios were

generated numerically using a load flow program and
stored in data files. Random errors were added to the
generated data to simulate typical measurement errors.
The added errors were uniformly distributed with a
specified range (for this data a standard deviation of 0.5%
was used for flow measurements and 0.02 degrees for
phase of synchronized measurements). Subsequently the
estimator was executed with the numerically generated
Figure 3. Example of Visualization – Voltage measurement data in order to evaluate its performance. In
Magnitude (tubes) and Phase (pies) Errors all tested scenarios the estimator converged within two to
four iterations with excellent results.

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 5
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

Kentucky

Section 1 Section 2 Section 3 Secti

G
PARADISE
G

G G
SHAWNEE
MARSHALL S UB

POCKET

MONTGOMRY
SU B

S UB
NAGEL
SU B
PHIPSBEND
G G
CUMBERLND SULLIVAN
S UB

S UB
GLEASON
WEAKLEY
WILSON
G G

S UB

SU B
VOLUNTEER
S UB

DAVIDSON
PINHOOK
SU B

BULL-RUN
G G

G G

LAGOONCRK JOHNSONVL ROANE

G G
SU B

Tennessee
SU B

MAURY
JACKSON S UB
G WATTS-BAR
G

ARKANSAS G
HAYWOOD
G

SHELBY
SU B S UB

FRANKLIN
SEQUOYAH
CORDOVA S UB Section 5 Section 6 G G

Section 7
BIRMSTEEL
G G

FREEPORT
G G

S UB

BENTON WIDOWSCRK
RACOONMT
G G

G G

SHAVEN MADISON G G
S UB
G G

BROWNSFER
BRFTRIC2
TRICO CONASAUGA
TRINITY SU B SU B

G G PLEASANTH BRFTRIC1
S UB

UNION BFTRIN2
BFTRIN1

EASTPOINT
SU B

SU B

LIMESTONE

CALEDONIA
LOWNDES
WVERNON
S UB

WESTPOINT
S UB G G

LOANOAK MILLER

Figure 4. TVA’s 500 kV System

Figure 7. Residuals of Bus Voltage Magnitude and
Phase – Scenario 1, Phase C – Magnitude
Magnification: Magnitude: 10, Phase: 100

The results of scenario 1 are illustrated in Figures 5, 6 and

7. Note that the errors are uniformly distributed and of
very low magnitude for all phases A, B and C. The range
of the errors is reported at the upper left corner of the
display.

The results of scenario 2 are illustrated in Figures 8, 9 and

10. Note that the errors for phase A are uniformly
distributed and of relatively low magnitude. The errors for
phases B and C (Figures 9 and 10) are substantially
greater. The source of these errors is the asymmetry of the
system.
Figure 5. Residuals of Bus Voltage Magnitude and
Phase – Scenario 1, Phase A – Magnitude
Magnification: Magnitude: 10, Phase: 100

Figure 8. Residuals of Bus Voltage Magnitude and

Phase – Scenario 2, Phase A – Magnitude
Magnification: Magnitude: 10, Phase: 100
Figure 6. Residuals of Bus Voltage Magnitude and
Phase – Scenario 1, Phase B – Magnitude Figure 11 illustrates the estimation results for the same
Magnification: Magnitude: 10, Phase: 100 measurement data set, but with a fictitious error (100
MW) introduced at one circuit MW flow measurements.
The visualization display shows both magnitude and
phase errors. Note that at one location of the network,

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 6
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

both magnitude and phase errors are much higher than

anywhere else. These increased errors clearly identify the
two end of the circuit with the “bad data”. It is important
to note that 100 MW for a 500 kV circuit is not really a
large error. Yet the three phase state estimator with the
visualization of Figure 11 clearly identifies the location of
the bad datum.

6. Summary and Conclusions

State estimators have a plethora of information that may
be overwhelming. The information include in a state
estimator solution may include bad data, overall
performance of the system, accuracy of the real time
Figure 9. Residuals of Bus Voltage Magnitude and model, etc. Visualization methods provide an excellent
Phase – Scenario 2, Phase B – Magnitude tool to communicate this information in a very effective
Magnification: Magnitude: 10, Phase: 100 way. The paper has discussed the state of the art of state
estimators, the existing biases and presented a hybrid state
estimator that is based on a full three phase model of the
system. Visualization methods provide a variety of state
estimator performance quantities in 3-D or 2-D displays.
The effectiveness of the visualization methods has been
demonstrated with three simple scenarios. These scenarios
illustrate the effectiveness of the visualization methods to
communicate the performance of the system, the expected
errors as well as any bad data. With respect to bad data
identification with visualization methods it is important to
note that the state estimator used in the research is a
hybrid estimator that uses highly redundant
measurements. In this case, bad data are immediately
identified with visualization methods. In case of low
redundancy it is possible that bad data become leverage
points which limit the effectiveness of the visualization
methods.
Figure 10. Residuals of Bus Voltage Magnitude and
Phase – Scenario 2, Phase C – Magnitude
Magnification: Magnitude: 10, Phase: 100
7. References
[1] Fred C. Schweppe and J. Wildes, "Power system static-state
estimation, Part I, II, and III" IEEE Transactions Power App.
Syst., vol. PAS-89, No.1, pp.120-135, January 1970.
[2] S. Zelingher, G.I. Stillmann, A. P. Sakis Meliopoulos,
"Transmission System Harmonic Measurement System: A
Feasibility Study," Proceedings of the Fourth International
Conference on Harmonics in Power Systems (ICHPS IV), pp.
436-444, Budapest, Hungary. October 1990.
[3] A. P. Sakis Meliopoulos, F. Zhang, and S. Zelingher,
"Hardware and Software Requirements for a Transmission
System Harmonic Measurement System," Proceedings of the
Fifth International Conference on Harmonics in Power Systems
(ICHPS V), pp. 330-338, Atlanta, GA. September 1992.
[4] A. P. Meliopoulos, F. Zhang, S. Zelingher, G. Stillmam, G. J.
Cokkinides, L. Coffeen, R. Burnett, J. McBride, 'Transmission
Figure 11. Residuals of Bus Voltage Magnitude and Level Instrument Transformers and Transient Event Recorders
Phase – Scenario 3, Phase B – Magnitude Characterization for Harmonic Measurements,' IEEE
Magnification: Magnitude: 10, Phase: 100 Transactions on Power Delivery, Vol 8, No. 3, pp 1507-1517,
July 1993.

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 7
 Proceedings of the 38th Hawaii International Conference on System Sciences - 2005

[5] A. G. Phadke, J. S. Thorp and K. J. Karimi, "State estimation Faculty of Electrical Engineering, Georgia Institute of
with phasor measurements," IEEE Transactions on Power Technology, where he is presently a professor. He is
Systems, Vol. PWRS-1, No.1, pp. 233-241, February 1986. active in teaching and research in the general areas of
[6] A. P. Sakis Meliopoulos, G. C. Cokkinides, and R. P. Webb, modeling, analysis, and control of power systems. He has
"Multiphase Power Flow Analysis," Proceedings of made significant contributions to power system
Southeastcon, Destin, Florida, pp. 270-275, April 4-7, 1982. grounding, harmonics, and reliability assessment of power
[7] K. A. Clements, O. J. Denison, and R. J. Ringlee, "The systems. He is the author of the books, Power Systems
Effects of Measurement Non-Simultaneity, Bias, and Parameter Grounding and Transients, Marcel Dekker, June 1988,
Uncertainty on Power System State Estimation", 1973 PICA Ligthning and Overvoltage Protection, Section 27,
Conference Proceedings, pp. 327-331, June 1973. Standard Handbook for Electrical Engineers, McGraw
Hill, 1993, and the monograph, Numerical Solution
[8] O. Alsac, N. Vempati, B. Stott and A. Monticelli,
“Generalized State Estimation”, IEEE Transactions on Power
Methods of Algebraic Equations, EPRI monograph series.
Systems, Vol. 13, No. 3, pp 1069-1075, August 1998. Dr. Meliopoulos is a member of the Hellenic Society of
Professional Engineering and the Sigma Xi.
[9] A. P. Sakis Meliopoulos, F. Zhang, and S. Zelingher, 'Power George Cokkinides (M '85) was born in Athens, Greece,
System Harmonic State Estimation,' IEEE Transactions on in 1955. He obtained the B.S., M.S., and Ph.D. degrees at
Power Systems, Vol 9, No. 3, pp 1701-1709, July 1994.
the Georgia Institute of Technology in 1978, 1980, and
[10] A. P. Sakis Meliopoulos and Fan Zhang, ‘Multiphase Power 1985, respectively. From 1983 to 1985, he was a research
Flow and State Estimation for Power Distribution Systems,' engineer at the Georgia Tech Research Institute. From
IEEE Transactions on Power Systems, Vol. 11, No. 2, pp. 939- 1985 to 2000, he has been with the University of South
946, May 1996. Carolina, Department of Electrical Engineering. Since
[11] B. Fardanesh, S. Zelingher, A. P. Sakis Meliopoulos, G. 2000 he has been a visiting professor of Electrical and
Cokkinides and Jim Ingleson, ‘Multifunctional Synchronized Computer Engineering at Georgia Tech. His research
Measurement Network’, IEEE Computer Applications in Power, interests include power system modeling and simulation,
Volume 11, Number 1, pp 26-30, January 1998. power electronics applications, power system harmonics,
[12] Sakis Meliopoulos, “State Estimation for Mega RTOs”, and measurement instrumentation. Dr. Cokkinides is a
Proceedings of the 2002 IEEE/PES Winter Meeting, New York, member of the IEEE/PES, and the Sigma Xi.
NY, Jan 28-31, 2002. Michael R. Ingram, P.E.. - is the Program Manager of
[13] T. J. Overbye, D. A. Wiegmann, A. M. Rich, Y. Sun,
Transmission Performance and Transmission Asset
"Human factors aspects of power system voltage contour Utilization Technologies with Tennessee Valley Authority
visualizations," IEEE Trans. on Power Systems, pp. 76-82, (TVA) in Chattanooga, Tennessee. He is responsible for
February 2003. the development and demonstration of new technologies
which improve electrical quality and reliability, increase
[14] A. P. Sakis Meliopoulos, George. J. Cokkinides and
Thomas J. Overbye, “Component Monitoring and Dynamic
power flow, and reduce operating expense of the TVA
Loading Visualization from Real Time Power Flow Model transmission system.
Data”, Proceedings of the 37st Annual Hawaii International Sandra C. Bell is a Project Engineer in the Power
Conference on System Sciences, p. 58 (pp. 1-6), Big Island, Delivery Technologies Group with the Tennessee Valley
Hawaii, January 5-8, 2004. Authority in Chattanooga, Tennessee. She is responsible
for supporting research, development and demonstration
[15] Mike Ingram, Sandra Bell, Sherica Matthews, A. P. Sakis
Meliopoulos and G. J. Cokkinides, ”Use of Phasor of new technologies which improve electrical power
Measurements, SCADA and IED Data to Improve State quality and reliability, increase power flow, and reduce
Estimation Procedures”, Proceedings of the 2004 Georgia Tech operating expense of the TVA transmission system.
Fault and Disturbance Analysis Conference, Atlanta, Georgia, Sherica A. Matthews is currently pursuing her Master's
April 26-27, 2004. in Electrical Engineering at the University of Tennessee,
Knoxville. Her research is in the areas of Power Systems
and Power Electronics. She is currently a Project Engineer
in the Power Delivery Technologies group with the
8. Biographies Tennessee Valley Authority in Chattanooga, TN. She is
testing and modeling ultra-capacitors for applications in
A. P. Sakis Meliopoulos (M '76, SM '83, F '93) was born high voltage transmission systems.
in Katerini, Greece, in 1949. He received the M.E. and
E.E. diploma from the National Technical University of
Athens, Greece, in 1972; the M.S.E.E. and Ph.D. degrees
from the Georgia Institute of Technology in 1974 and
1976, respectively. In 1971, he worked for Western
Electric in Atlanta, Georgia. In 1976, he joined the

Authorized licensed use limited to: Centro De Pesquisas De Energia Eletrica Cepel. Downloaded on May 30,2023 at 17:55:42 UTC from IEEE Xplore. Restrictions apply.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE 8