Documente Academic
Documente Profesional
Documente Cultură
1, FEBRUARY 2001 87
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
88 IEEE TRANSACTIONS ON EDUCATION, VOL. 44, NO. 1, FEBRUARY 2001
II. VECTOR NOTATION FOR CIRCUIT EQUATIONS Finally, two different SIMULINK blocks representing trans-
mission line models are given.
In order to build models for the generator, load, and transmis-
sion line, it is first necessary to convert the basic circuit equa-
A. Model of a Generator
tions from phasor notation to a two-dimensional vector nota-
tion, since SIMULINK does not accept complex-valued states. The simplest generator model is a slack bus, shown in Fig. 2.
Throughout the paper, we will use the following notation to rep- It consists of an independent voltage source and an internal
resent vectors and matrices: impedance . This impedance can also be taken to be zero. The
block is designed by viewing a slack bus as shown in Fig. 2.
(1) The matrix equation for the slack bus is
(11)
(2)
The “swing” type generator model [4], [5] is the simplest
model capturing only the real power dynamics and assuming
(3) constant voltage magnitude. However, the angle of the voltage
source is a state variable, , which changes according to the fol-
(4) lowing inertial equation:
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
ALLEN et al.: INTERACTIVE OBJECT-ORIENTED SIMULATION OF INTERCONNECTED POWER SYSTEMS USING SIMULINK 89
Fig. 5. State-space subblock representing dynamics of the IEEE Type 1 exciter in the swing-type generator.
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
90 IEEE TRANSACTIONS ON EDUCATION, VOL. 44, NO. 1, FEBRUARY 2001
(22)
B. Model of a Load capacitance can also be included, if needed. The state equations
The next step is to develop a model of a load. Two static for the line are
load models are used; constant impedance and constant power. (29)
A diagram applicable for both load models is shown in Fig. 6. In
the constant impedance model, is obviously treated as a fixed
quantity; therefore, the constant impedance load is represented (30)
by the equation Note that .
A dynamic model is also included for a transmission line with
(23) series capacitive compensation. This model has four states and
evolves according to [6], [7]
To simulate the load behavior of a constant power load, the
admittance of the load becomes a state variable with (31)
the following dynamics:
(32)
(24)
(33)
(25)
(34)
is a time constant (typically 0.1 s), while ( ) are the
desired constant real (reactive) power for the load. and are The derivation of this model is also based on time-varying pha-
the instantaneous real and reactive power at the load. In these sors; the details can be found in [6] and [7].
equations, a positive implies power supplied to the network; To demonstrate relevance of representing transmission lines
therefore, for a load, will be negative. using recently developed time-varying phasor algebra, the dif-
ference between the two responses obtained by means of two
C. Model of a Transmission Line models, i.e., an algebraic model and the model based on time-
The transmission lines are usually modeled as -sections con- varying phasor algebra, is shown in Fig. 8 for the system shown
sisting of resistors, inductors, and capacitors that have the fol- in Fig. 9. The more theoretically sound dynamic model allows
lowing voltage–current relations [5], [9], [10]: for detecting unstable subsynchronous dynamics, while simula-
tions using an algebraic model fail to detect this type of insta-
(26) bility. This is just one example of using SIMULINK blocks of
various accuracy for simulating dynamic phenomena of interest.
(27)
D. Interconnected System
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
ALLEN et al.: INTERACTIVE OBJECT-ORIENTED SIMULATION OF INTERCONNECTED POWER SYSTEMS USING SIMULINK 91
(35)
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
92 IEEE TRANSACTIONS ON EDUCATION, VOL. 44, NO. 1, FEBRUARY 2001
Fig. 10. The three bus system and its SIMULINK block diagram.
Complexity is managed by storing machine data in matrices This approach serves the dual purpose of making access to the
that are serviced by a number of special-purpose routines. Each data convenient, while isolating the raw data structure from both
subsystem block is indexed by two identifiers, one of which the user and the simulation model. New model types are accom-
indicates its type (round-rotor gepnerator, static VAR compen- modated by assigning an unused system type identifier to the
sator, etc.) and one which locates its parameters in the data array. block, defining its particular raw data structure and giving it a
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
ALLEN et al.: INTERACTIVE OBJECT-ORIENTED SIMULATION OF INTERCONNECTED POWER SYSTEMS USING SIMULINK 93
Fig. 12. Comparison of system response for two different load types.
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
94 IEEE TRANSACTIONS ON EDUCATION, VOL. 44, NO. 1, FEBRUARY 2001
slot in the data array. Once this is done, the model can be used of the generators, allowing its representation as a set of al-
and its parameters set and retrieved at will without knowledge gebraic relations. In order to solve the network equations in
of the actual position or structure of the data record. one step rather than iteratively, the modularity of the system
Routines are available that then initialize generic system is compromised to the extent that the machine impedances
blocks with the appropriate data, based on the machine iden- are made availaPble to the routine that calculates the network
tifier, which is the only quantity that must be entered by the currents.
user in the block mask. In particular, depending upon the level of detail of the
In this way modularity is preserved to a great extent, generators, if transient or subtransient saliency is ignored and
however we do allow some sacrifice of subsystem data loads are modeled as constant impedances, it is possible to
localization in order to facilitate the solution of the intercon- form the reduced admittance matrix and solve for the currents
nection constraints. Because of the need to limit the number in a single matrix multiplication [5]. At present the simulation
of dynamic variables in larger simulations, the transmission of a particular fault scenario is accomplished by calculating
network is modeled in the more traditional manner, i.e., its separate reduced admittance matrices for the prefault, faulted
dynamic behavior is considered to be much faster than that and postfault networks. These are then loaded at the appro-
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.
ALLEN et al.: INTERACTIVE OBJECT-ORIENTED SIMULATION OF INTERCONNECTED POWER SYSTEMS USING SIMULINK 95
priate time by a MATLAB routine that reads the system time Eric Allen received the B.S. degree in electrical engineering from Worcester
and sequences the fault scenario. Polytechnic Institute, Worcester, MA, in 1993. He received the S.M. degree in
electrical engineering from the Massachusetts Institute of Technology (MIT),
As a final consideration, we note that it is important for the Cambridge, in 1995, studying the effects of torsional shaft dynamics on non-
subsystem blocks to be initialized at the system equilibrium. linear generator excitation control. He received the Ph.D. degree in electrical
Therefore routines are included in each subsystem block that engineering from MIT in 1998, with the dissertation titled “Stochastic unit com-
mitment in a deregulated electric utility industry.”
calculate the equilibrium state for that block, based on the initial His research interests include dynamic systems and control in electric power
load-flow data. This also facilitates the recovery of linearized systems, dynamic programming for commitment decisions in electricity mar-
system matrices via the “linmod” function. The size of the larger kets, and stochastic control.
models and the existence of multiple equilibria present obstacles
to the successful determination of the system equilibrium via
the MATLAB routine “trim,” but “trim” can be useful for cal-
culating perturbed equilibria, using the initial state as a starting Niels LaWhite received the S.B. degree in 1987 and the S.M. and E.E. degrees
point. SIMULINK was shown to be a useful tool for simulating in 1995, all in electrical engineering, from the Massachusetts Institute of Tech-
an aggregate model of the NPCC system by representing it as a nology (MIT), Cambridge.
He worked in the wing power industry, designing instrumentation and control
system with 29 machines [12]. systems for large scale wind turbine generators. Recently, he has been working
as a Consultant and conducting research in digital signal processing algorithms
VI. CONCLUSION for measuring power and other electrical quantities. His other research interests
include wind turbine control, modeling, and vibration analysis.
In conclusion, a variety of simulations in this paper show
SIMULINK-based time domain responses of an interconnected
system with different controls present on the system. The flex-
ibility of SIMULINK as a very useful tool for f simulations of
Yong Yoon received S.B. degrees in applied mathematics and in electrical engi-
small power systems, intended to analyze specific phenomenon neering and the M. Eng. degree in electrical engineering from the Massachusetts
has been demonstrated in this paper. Institute of Technology (MIT), Cambridge, in 1995 and in 1997, respectively.
For further information, the interested reader may contact He is currently pursuing the Ph.D. degree in electrical engineering at MIT with
the concentration on electric power system economics engineering.
Marija Ilic at ilic@mit.edu via e-mail. His research interests include modeling of energy markets as stochastic dy-
namic systems, developing concepts for the Independent Transmission Com-
REFERENCES pany and designing software tools for various energy market participants. He
has strong backgrounds in control, estimation, mathematics, research design,
[1] SIMULINK: Dynamic System Simulation Software. Natick, MA: The and regulatory economics.
Mathworks Inc., 1994.
[2] G. Rogers and J. Chow, “Hands-on teaching of power system dynamics,”
IEEE Comput. Applicat. Power, pp. 12–16, Jan. 1995.
[3] N. LaWhite and M. Ilić, “Vector space decomposition of reactive power
for periodic nonsinusoidal signals,” IEEE Trans. Circuits Syst. I, vol. 44,
Apr. 1997. Jeffrey Chapman received the B.S. degree from the University of California,
[4] P. W. Sauer and M. A. Pai, “Modeling and simulation of dynamic sys- Santa Barbara, and the M.S. and Ph.D degrees, all in electrical engineering, from
tems,” in Advances in Control and Dynamic Systems. New York: Aca- the Massachusetts Institute of Technology (MIT), Cambridge, in 1990, 1992,
demic, 1991, vol. 43. and 1996, respectively.
[5] M. Ilić and J. Zaborszky, Dynamics and Control of the Large Electric He is currently the West Coast Manager for Whole Sale Trading Applications
Power Systems. New York: Wiley, 2000. with ABB Energy Information Systems. His interests include dynamics of elec-
[6] E. H. Allen, J. W. Chapman, and M. Ilić, “Eigenvalue analysis of the tric power systems and the electricity markets.
stabilizing effects of feedback linearizing control on sub-synchronous
resonance,” in Proc. 4th IEEE Conf. Contr. Applicat., Schenectady, NY,
September 1995.
[7] E. H. Allen, “Effects of torsional dynamics on nonlinear generator con-
trol,” Master of Science thesis in electrical engineering, Massachusetts
Inst. Technol., Cambridge, Feb. 1995. Marija Ilić (S’78–M’80–SM’86–F’99) received the Dipl. Eng., M.Sc., M. Eng.,
[8] M. Ćalović, Dynamic State-Space Models of Electric Power Systems: and D. Sc. degrees from Washington University, St. Louis, MO, in 1980.
Univ. Illinois, Urbana-Champaign, 1971. She was an Assistant Professor at Cornell University, Ithaca, NY, and
[9] C. L. DeMarco and G. C. Verghese, “Bringing Phasor Dynamics Into tenured Associate Professor at the University of Illinois at Urbana-Champaign.
The Power System Load Flow,” Univ. Wisconsin-Madison, EECE Dept., She is currently a Senior Research Scientist in the Department of Electrical
Tech. Rep. ECE-93-10. Engineering and Computer Science at the Massachusetts Institute of Tech-
[10] J. Zaborszky, H. Schattler, and V. Venkatasubramaninan, “Error estima- nology (MIT), where she teaches several graduate courses in the area of electric
tion and limitation of quasistationary phasor dynamics,” in Proc. Power power systems and heads research in the same area. She has over 20 years
Syst. Comput. Conf., Avignon, France, Aug. 1993, pp. 721–729. of experience in teaching and doing research in this area. She is coauthor of
[11] M. Ilić and X. Liu, “A simple structural approach to modeling and anal- the books entitled Dynamics and Control of Large Electric Power Systems,
ysis of the inter-area dynamics of the large electric power systems,” in (New York: Wiley, 2000) and Hierarchical Power Systems Control: Its Value
Proc. 1993 North Amer. Power Symp., Washington, DC, October 11–12, in a Changing Industry (London, U.K.: Springer-Verlag, 1996), and coeditor
1993, pp. 560–578. of the book entitled Electric Power Systems Restructuring: Engineering and
[12] J. W. Chapman et al., “Stabilizing a multimachine power system via Economics (Norwell, MA: Kluwer, 1998). Her main interests include the
decentralized feedback linearizing control,” IEEE Trans. Power Syst., systems aspects of operations, planning, and economics of electric power
vol. 8, pp. 830–839, Aug. 1993. industry.
Authorized licensed use limited to: Reva Institute of Tehnology and Management. Downloaded on October 11, 2008 at 04:42 from IEEE Xplore. Restrictions apply.