1

Study on Wide Area Measurement System based

Transient Stability Control for Power System
F. F. Song, T.S. Bi, Member, IEEE, and Q.X. Yang

Abstract--In this paper, the state of the arts of power systems available. Up to now, over 50 PMUs have been equipped in
in China and some applications in transient stability control Chinese power systems. From 2002, real time monitoring
based on wide area measurement system (WAMS) are introduced systems based on PMU have been established in succession in
briefly. With the development of power grid interconnection and
the power systems of South China, West China, East China,
transmission utility deregulation, the operation of electric power
system become more and more complex, which challenges the Sichuan Province, three-gorge (left bank), north China and
security and transient stability control of power systems. The northeast China. Now, a new information platform is
advent of WAMS opens a new path to the system trajectory generated, which consists of WAMS, EMS/SCADA,
prediction and real time closed-loop stability control to prevent information of protection system, data of stability control
power system outage. The defects of current transient stability system and part of prior off-line knowledge. This makes it
control system are analyzed, then a possible scheme of transient
possible to predict the system trajectories and realize real time
stability control based on WAMS is proposed, and the
corresponding future work is put forward as well. Finally, a new closed-loop stability control to prevent power system
way of solving cascading events is discussed. blackout.
In this paper the state of the arts of power system transient
Index Terms--Transient stability, control, wide area stability control is introduced briefly, and some defects of
measurement system, power systems. current power grid control are described in detail in section 2.
Section 3 proposes a possible scheme of transient stability
I. INTRODUCTION control and indicates that the design may improve power

O peration of electric power system becomes more and transmission limit and avoid operator misoperation. Finally
more complex and brings challenges to the security since some discussions on cascading events are given.
grid has been in a era of bulk power system in the world.
So many new components such as HVDC, power electronic II. THE OVERVIEW OF CURRENT STABILITY CONTROL IN
devices and distributed generators have been commissioned in POWER SYSTEMS
the power grid. In China, up to 2020, the generation capacity The existing emergency stability control devices are
of Chinese power grid will be 900 million kW and the power
distributed in some venerable sites. These sites are chosen by
grid will possess more than 20 high voltage DC lines. Besides,
prior off-line computation based on N-1 stability analysis.
the power transmitted from west to east will be over 100
They are inter-independent and entirely serve for the local site.
million kW. Meanwhile, under the deregulated market
conditions, the transmission utilities are finding it more Then a strategy table for stability control under potential
difficult to upgrade the transmission systems as in the past and contingencies is worked out by off-line simulation.
unable to keep up with the growing loads and the added Once there is a disturbance, the first step is to identify the
generation. The customers try to buy cheaper electricity which stability of the system. When the system might be unstable,
might be generated in a remote area and make the the control devices start searching the control strategy table
transmission distance longer and transmission network more based on the operation mode of pre-fault power system and
complex. The super-grid has also become so venerable that the the disturbance. Then they find the corresponding control
events in one area can cascade and have significant impact on measures and quantities, which will be executed quickly to
the remote areas. All these changes are putting pressure for the prevent blackout. But there are two problems: one is that
development of new tools to monitor and analyze system because the operation modes and power network topologies
operation and take measures to realize transient stability are time-varying and thus the strategy table cannot cover all
control if necessary. the contingency cases. The other is that because the strategy
Wide area measurement system (WAMS) opens a new path table is established by simulating computation, which is
to power system stability analysis and control. The snapshot of dependent on models and parameters of power systems. But
the system dynamic behavior can be obtained and updated the accurate models and parameters of power system are hard
every 20 ms by WAMS, i.e. the historical data of phasors are to achieve. For example the outage of America cannot be
recurred by simulation computing based on original models
F.F. Song, T.S. Bi and Q.X. Yang are all with Si Fang Research Institute, and parameters in 1996 [1].
North China Electric Power University; Key Laboratory of Power System Therefore, the operation limit is conservative in the current
Protection and Dynamic Security Monitoring and Control (North China
Electric Power University), Ministry of Education; Beijing 102206, P.R.
state. The local control devices and system control center
China (e-mail: [email protected], [email protected], and (EMS) are disjunct. The local control devices cannot globally
[email protected]).

Authorized licensed use limited to: Air University. Downloaded on July 17,2020 at 11:09:14 UTC from IEEE Xplore. Restrictions apply.
 2

control system by using local information. Besides, the quick in the control center, and then is sent to local control devices
local control and slow center control are not coordinate. The in real-time. Therefore, the power system can operate in a
existing stability control system is helpless to the cascading wider region without losing security and avoid wasting the
event which occurs in many outages as shown in Fig. 1-3. resources of electric power.

Fig. 1. A fault occurs on line 2

Fig. 2. The relay of line2 trips the corresponding breakers

Fig. 3. System B is collapsing

The whole system consists of the subsystems A, B and C.
Subsystem A and C supply power to subsystem B. There is
one tie-line named Line 1 as shown in Fig.1 between system A Fig. 4. The global stability control system
and B and two tie-lines named Line 2 and Line 3 between
When a disturbance occurs in the power system, in order to
system B and C. At first, there is a fault occurs on Line 2 realize the real-time stability control in power system, it is
which may be the result of a branch of the tree reaching the necessary to predict the trajectories of the system and decide
Line 2. Then the relay of Line2 trips the corresponding whether the system is stable or not first. Some prediction
breakers, which leads to the power flow of Line 2 transferring methods based on WAMS are proposed in recent years. For
to Line 1 and Line 3. After a time-delay, the overload relays of example: curve extrapolated method, intelligent prediction
Line 1 and Line 3 act and trip the corresponding breakers in method and some other methods base on models of power
Line 1 and Line 3. Finally, the system B is isolated. A system. Curve extrapolated methods such as polynomial curve
blackout will occur in system B, if there are not enough fitting method [2]-[4], self-regression method [5] and some
backup power supplies in it. Meanwhile, power flow between intelligent prediction methods [6]-[11], have been proposed.
supplier and consumer are unbalanced in the system A and C However the prediction accuracy cannot be guaranteed. In
and their stability control devices take measures to solve the [12]-[13], there is another method to realize fast prediction
problem individually. It can be seen that all relays act based on simple system model using PMU measurements. But
correctly as expected in the whole process, but the outage this method needs several hypotheses, which limit its
occurs in the system B. The possible reasons of this cascading applications. We are trying to study on trajectories prediction
event are that the relays and stability control devices only based on a curve fitting technique described in another paper.
protect their own parts, and are lack of the global concept. So, When the control center finds that the system will be
a global stability control system is proposed as described in unstable after running the prediction algorithm, it will make
some control strategies on-line according to the predicted
section 3.
trend of the power system and send these strategies to area
control stations. These strategies will be translated and then
III. A SCHEME OF GLOBAL STABILITY CONTROL SYSTEM
sent to the corresponding local control devices to execute in
The block diagram of the global stability control system is the area control station. Therefore the control strategies are
proposed as shown in Fig. 4. The control center receives done before the power system is deteriorating. If the operation
global system data from WAMS, EMS/SCADA, protection state of the system still remains unstable, the control center
and area stability control to analyze system stability and will refresh the control strategies again and again till the
compute the control strategies, and then sends control system returns to a stable state. This is the whole closed-loop
commands to area control station. The area control stations are control flow chart as shown in Fig. 5.
in charge of gathering and coordinating data from local control Fig. 5 shows that there is a feed back loop in the algorithm.
devices and send them to the control center and send the This structure ensures the stability control strategies to be
commands from control center to local control devices. The updated in real-time through feed back control.
local control devices collect data from power system and send This design use dynamic information of whole system, it
them to area control station and execute the commands from can solve the problem of stability in a global way and also can
the control center. The control center is the main part of the coordinate with original local control devices. If the problem
is in the local system, the corresponding local control devices
global stability control system and making control strategies is
can act quickly before the global control act, then this act
the main function of the control center.
information will be sent to the control center and the control
When operation of power system is in a normal condition, center will analyze the next step. When the control strategies
with the variety of operation modes and network topologies in local control devices collide with these in global control
and system parameters, the transmission limit can be updated

Authorized licensed use limited to: Air University. Downloaded on July 17,2020 at 11:09:14 UTC from IEEE Xplore. Restrictions apply.
 3

system, local control should comply with the strategies in 5. Study on assessing methods and index of the validity of
global control system. models and parameters in power system simulation.
The solutions of all above mentioned problems and the
corresponding numerical examples would be published in
papers later.

IV. STRATEGY TO PREVENT CASCADING EVENT

Current stability control devices cannot solve the problem
of the cascading events as shown in Fig. 1-3, but the global
stability control system can obtain the whole information of
the system such as WAMS, EMS/SCADA, protection and
stability control devices, so, it is possible to solve the
problems as described in detail as bellows.
Because the control center can receive data from WAMS,
EMS/SCADA, protection and control devices, the action
signal of relays and trip signal of breakers in Line 2 are
received by the control center. Then the case of resulting
power flow transfer in Line 1 and 3 is analyzed before their
overload or backup protections acting in the control center.
Fig. 5. The flow chart of stability control algorithms in control center If the case is only a power flow transfer, then the
corresponding control strategies are made in control center
No matter in the stability prediction or control, the according to the balance of supply and demand. The control
simulating computations are necessary. So the accuracy of center will send power generation control, power generation
simulating computations should be assured, i.e. the system and load control, power generation control to system A, B, and
models and parameters should be validated first [14]. But it is C respectively as shown in Fig. 7. i.e. according to the balance
impossible to do the experiments arbitrarily in power systems. of supply and demand in system A, B, and C, some generators
Fig. 4 show the historical dynamic process of the global power should be tripped in system A and C, while some generators
system is captured by the control center. So, the possible path should increase their output and some loads should be shed in
of model and parameter validation can be done through these system B. Therefore, the power flow is decreased in Line 1
dynamic data as shown in Fig. 6. and Line 3, the corresponding relay can not act any more and
power supply form A and C to B is remained.
If the case is a fault occurring in Line 1 or Line 3 also, the
corresponding main protections will act and trip its breakers.
The action time of main protection is about a few
ten-milliseconds, the control center receives the protection
action signal from Line 1or Line 3 when it is analyzing the
power flow increase in Line 1 or Line 3 during this time, so
the analyzing is refreshed by added the new information in the
Fig. 6. Schematic diagram of simulation control center.

In Fig. 6, Δ represents the validity of models and parameters

in power system. It can be used as assessing the validity of the
models and parameters in power system.
After simulation assessed by the field data several times, the
relation between the accuracy of simulation and validity of
models and parameters will be found. Therefore, although the
prediction result by simulation is not verified by actual system
data, the accuracy and certainty of it can be assessed too.
Finally, there are following problems to be studied in the
future in the whole scheme of global stability control system.
1. Study on a rule of dynamic setting for the transmission Fig. 7. A scheme of stability control
limits, and in which the setting can be computed quickly. A problem needed to be studied in the future is that if a fault
2. Study on new trajectory prediction methods based on occurs in a line and its main protection refuses to act and then
WAMS and identification theories and algorithms of system the line is overloading, how to decide the reason is power flow
stability. transferring of other lines or a fault occurring in it.
3. Study of some data digging techniques and information
processing methods, V. CONCLUSIONS
4. Study on global stability control algorithms based on
WAMS, and how to make the algorithms when time delaying The overall design of a global control system is proposed in
is considered. this paper, its functions including dynamic revising setting of

Authorized licensed use limited to: Air University. Downloaded on July 17,2020 at 11:09:14 UTC from IEEE Xplore. Restrictions apply.
 4

transmission limits, perturbed trajectory prediction and [8] F. Lin, Y.T. Liu, and X.Z. Qiu, "Power System Transient Stability
Decision Making Basedon Fuzzy Neural Networks," Proceedings of
stability identification and global stability control. This design China Electrotechnical Society, vol. 16, pp.83-87, Apr. 2001.
is the combination of electric engineering and communication [9] Q. Zhang, Z.X. Han, S.K.Tso, and X.P. Gu, "Input Dimension
technologies. It is based on the global information and aims at Reduction In Neural Network Training for Transient Stability
the security and reliability of power system operation. The Assessment, " Automation of Electric Power Systems, vol. 25, pp.32-35,
Feb. 2001.
suggested scheme can revise setting of transmission limits [10] T.F.Robert, K. Ah and C.S. Harry, "Rughooputh. Real-time Transient
flexibly according to the operation state in power system. Stability Prediction Using Neural Tree Network," In Proc. Of IEEE
With perturbed trajectory prediction and stability International Conference, vol. 3. 22-25 pp. 2182- 2187, Oct. 1995.
identification, it can estimate the trend of system operation, [11] S. Rovnyak, S. Kretsinger, J. Thorp and D. Brown, "Decision Trees for
Real-time Transient Stability Prediction," IEEE Trans. on Power
and so that it can win time for preparing control strategy Systems, vol. 9, pp.1417-1426, Aug. 1994.
before the system deteriorate. Moreover, the future works in [12] C.W. Liu, and J.S. Thorp, New Methods for Computing Power System
the suggested scheme are summarized. Model and parameter Dynamic Response for Real-time Transient Stability Prediction. IEEE
validation is essential for trajectory prediction and control. So, Transactions on Circuits and Systems- Ⅰ : Fundamental Theory and
a new idea of model and parameter validation based on Applications, vol. 47, pp.324-337, Mar. 2000.
[13] J.N. Peng, Y.Z. Sun, and H.F. Wang, "Research on the Perturbed
WAMS is proposed. The global control system is under Trajectories Prediction Based on Wide Area Measurement and On-line
developing and the simulation results would be published Admitance Matrix Identification," Automation of Electric Power
later. Systems, vol. 27, pp. 6-11, Nov. 2003.
[14] R.M. He, Power System Dynamics Monitoring and Management
System," Automation of Electric Power Systems, vol. 26, pp. 1-4, Mar.
2002.
VI. REFERENCES
[1] B.H. Zhang, "Strengthen the Protection Relay and Urgency Control VII. BIOGRAPHIES
Systems to Improve the Capability of Security in The Interconnection
Power Network," Proceedings of the CSEE, vol. 24, pp. 1-6, Jul. 2004.
F.F. Song was born in China, on Jun 15, 1977. She received her BS and MS
[2] Z.L. Lu, B.H. Zhang, and H.X. Ha, "Real-time transient stability
degrees in power system & Automation Department from North China electric
prediction for muti-machine power system based on phasor
power university in Baoding in 1999 and 2003 respectively. Now she is a
measurement units," Relay, vol. 28, pp. 3-5, Jan. 2000.
Ph.D. candidate in North China Electric Power University in Beijing. Her
[3] F. Lin, W. Zhang, and Y.T. Liu, "PMU Based Transient Stability
interests are power system stability prediction and control based on WAMs.
Predication in Power System," Relay, vol. 28, pp. 33-35, Nov. 2000.
[4] J.S. Su, and C. Chen, "Transient Stability Prediction Using Time-Series
Based on GPS Synchronized Measurement," Electric Power Automation T.S. Bi (St.M’1998, M’2002) received her B.Sc. (Eng.) and M.Sc. (Eng.)
Equipment, vol. 21, pp.7-9, Sep. 2001. degrees in EE from North China Electric Power University in 1994 and 1997
[5] Q. Guo, X.P. Liu, S.R. Lu, and D.Z. Xia, "Application of GPS respectively and her Ph.D. degree in EE from the University of Hong Kong in
Synchronized Clock to Power System Transient Stability Predict and 2002. Since then she has joined North China Electric Power University as an
Control," Automation of Electric Power Systems, vol. 22, pp.11-13, Jun. associate professor. Now she is the vice director of Key Laboratory of Power
1998. System Protection & Dynamic Monitoring and Control under Ministry of
[6] Y.T. Liu, and F. Lin, "Application of PMU and Fuzzy Radial Basis Education. Her research interest is power system control, protective relay and
Function Network to Power System Transient Stability Prediction, " the application of artificial intelligence in power systems.
Proceedings of the CSEE, vol. 20, pp. 19-23, Feb. 2000.
[7] J.S. Su, P.J. Liao, and D.M. Zhou, "Power Systems Transient Stability Q.X. Yang received his B. Sc. (Eng.) and M.Sc. (Eng.) degrees in EE from
Prediction by Using Fuzzy Neural Network Based on GPS Synchronized Zhejiang University in 1957 and 1960 respectively and his PhD. Degree in EE
Measurements," Relay, vol. 29, pp.13-15, Feb. 2001. from the University of New South Wales, Australia. Since 1960, he has joined
North China Electric Power University. His research interest is power system
control and protective relaying. He is a Chinese Academician of Engineering

Authorized licensed use limited to: Air University. Downloaded on July 17,2020 at 11:09:14 UTC from IEEE Xplore. Restrictions apply.