Beruflich Dokumente
Kultur Dokumente
Operations
Power system operators often reach a cognitive barrier when The picture is quite different, however, when we look at
information arrives too fast during a power system emergency.At the use of an EMS during an unforeseen event or a failure
such times it becomes difficult to reach a correct diagnosis of the
problem or to formulate the correct decision when actions must of major components on the power system. In such
be taken. Artificial Intelligence gives designersof Energy Manage- instances, the EMS serves mainly as an information gath-
ment Systems a way to solve many of the diagnosis and decision ering and reporting system and the sophisticated appti-
problems so as to make the EMS more useful. This paper explores cation s~ftware that functions in normal operation may be
reasons why AI techniques, such as knowledgebased expert sys- of little use. For example, in a regulatory shutdown of all
tems, arebeing used in EMS designs and the differences between
knowledgebased expert systems and traditional numeric a l p nuclear units, human operators will take overtheeconomic
rithm development. The differences between expert systems and dispatch and unit commitment to resc:hedule economic
the numeric approach extend to the basic conception and design operation. Similarly, when a sudden loss o f transmission
of the applications. This is illustrated using a relay fault diagnosis equipment occurs it is human operators who must under-
system, showing boththe traditional and rapid prototyping
approaches to its development. Finally, issues concerned with the stand what has happened and decide on what actions to
implementation of AI in EMS computers are explored along with take. It is especially during such emergenciesthat conven-
the authors' predictions of possible AI applications to power sys- tionalsoftwareis1esseffective.Therequirementforsmarter
tem operations. software thus becomes more important in such instances.
Coping with emergency events is reflerred to as a diag-
I. INTRODUCTION nosis and decision process.Suchprocessesare illstruc-
Modern power systems are operated by highly skilled tured and their solutionrests heavily on the experience and
operators through computerized control systems. The skill of the human operators to react correctly. The key to
energy managementsystem (EMS)is the center of a control human ability in such situations lies in the experience with
system organized in a hierarchical structure utilizing remote sirnilareventsandtheuseof heuristicstomapexistingcom-
terminal units, communication links, and various levelsof plex situations onto learned past eventsto solve problems.
computer processing systems. The function of the EMS is Since Artificial Intelligence allows the realization of heu-
to ensure the secure and economic operation of thepower ristic techniques in a computer, the waly is now open for
system as well as to facilitate the minute-by-minute tasks many new applications of computers in power system oper-
carried out by theoperations personnel. The EMS is mainly ations.
designed to be used in the"norma1"state where such func- There is also a need to incorporate heuristics in many of
tions as state estimation, security analysis, and optimal the functions now carried out by large application pro-
power flow are usedto ensure secureoperation while func- grams. For example, security analysis programs may fail to
tions such as automatic generation control, economic dis- converge or economic analysis programs may fail to meet
patch, unit commitment, and load forecasting are used to certain operating constraints because the constraints are
ensure that the most economic operation is obtained. Much difficult to express mathematically. In such cases, human
of what happensin normal operation is now computerized intervention must be relied on tosolve or circumvent the
and human operators only intervene to carry out the few problem. Research usingartificial intelligenceisalso begin-
manual tasks required. ning to yield methods of embedding complex heuristics
into conventional application software for use in normal
Manuscript received November 17,1986; revised May 10,1987. power system operations.
6. F. Wollenberg iswith Control Datacorporation, Minneapolis, In this paper we are specifically addressing the use of
MN 55441, USA. expert or knowledge-based systems wlhich is one of the
T. Sakaguchi i s with the
Central Research Laboratory,Mitsubishi principal branches of artificial intelligence. Several exper-
Electric Corporation, 1-1, Tsukaguchi Honrnachi 8 Chome, Arna-
gasaki, Hyoga 661, Japan. imental and near-practical-use expert systems have been
I E E E Log Number 8717877. developed for use in power system operations. Sakaguchi
o 1987 IEEE
oola~191~7/120(r16mw1.m
(p rule-name
(premise-part) + (action-part) )
SUBSTATION FRAME:
Name: character string
Breakers: (pointer to breaker frame)
Bus sections: (pointer to bus section frame)
Lines: (pointer to lineframe)
Transformers: (pointer to lineframe)
BREAKER FRAME:
Name: character string
Duty rating: real constant
Terminal bus sections: (pointers to bus section objects)
Status: openlclosedlunknown
determination: access-breakerdata
LINE FRAME:
Name: character string
Rating: real constant
Terminal bus sections: (pointers to bus section objects)
Status: connectedlopenendedlde-energized
determination: line-status-algorithm
1680 PROCEEDINGS OF THE IEEE, VOC. 75, NO. 12, DECEMBER 1987
This type of structure allows an expert system that knows The strength of AI techniques over conventional pro-
something (i.e., some object has been set to a value in the gramming can be summarized a s follows:
STM) about some part of thesubstation (say, the status of 1) flexibility: Expertsystemsare suited to solving ill-
a breaker) to use the frame relationships to directly infer structured problems. Furthermore, the environment used
conditions of other objects in the substation (say, the status to construct expert systems allows them to be prototyped
of a line). The Automatic ReasoningTool (ART) expert sys- rapidly and incrementally so that many alternatives and fre-
tem language [I41is an example of this type of structure. quent updatings can be tried.
2) HighPerformance: Expert systems try to implement the
C. Logic-Based System level of performanceexhibited bya personwith recognized
expertise in the problem domain.
The frameworks we havedealt with so far areappropriate 3) Understanding: Expert systems canexplain the lineof
to represent procedural knowledge such as: what to do reasoning used as well as the contents of the knowledge
when certain conditions are met. A different way t o r e p base. This is a key element when the designer debugs the
resent knowledge requires one to specify “what” instead system and increases the confidence of the user.
of “how.“ A logic-basedsystem provides us with such
means. Prolog i s a programming language to represent a IV. DEVELOPMENT
OF AN AI APPLICATION
“what”-type knowledge based on predicate calculus [15].
For example, the following Prolog statement might be The development of AI programs usually involves dif-
used in an expert system to guide operators in system res- ferent proceduresthan used in the development of numeric
toration: programs. To illustrate this difference we review an appli-
cation that deals with the diagnosis of power system faults
restoration-required( X, yes ) + and is constructed using both numeric and AI programs.
component( X ), The problem is to identify a faulty element in a power
charge-state( X, Y ), =I=( Y, is-charged), system by observing the relay and circuit breaker tripping
fault-state( X, Z ), =/=(Z, is-faulted). signals. Fig. 3 illustrates a roughoutline for a conventional
Fig. 4. Powersystemfaultproblem.
theassumptions, possible unoperated relays, and clearer. Thisproblem is made worsebythe fact that numeric
measurements until the designated element is programming languages are less readableand less under-
accepted or eliminated. If eliminated try another ele- standable when applied to knowledge-intensive problems.
ment. The experience in applying an expert system to the same
3) The third case involves no elements in the con- problem is quite different. The software development par-
junction and results from more than double ,faults adigm is illustrated in Fig. Sand iscalled the"rugbymode1."
within that area or some relay misoperations.Find a set
of active relays so that the set of protected elernents
has a nonempty conjunction. If successful, applry the
diagnosis algorithm to each set, otherwise assume relay
misoperation.
Now coding the algorithm is begun. Thealgorithm might
be written ina system description language andits source
code compiled into Fortran source code. This is followed
by the usual compile, load, link, and run sequence which
requires repeated changesto the program source code and
may even require changes to the system description itself.
The human resources to accomplish this task are esti-
mated to be eight man-monthsfor the first phase, four man- Fig. 5. The "rugby model" of AI software development.
monthsforthesecond,andsixman-monthsforthelast.The
total amounts to one and a half man-years. Thefinal Fortran As in the pipeline model shown in Fig. 3, there are design,
source code i s estimated to be about 15 thousand lines and coding, and debug phases. However, they are strongly
will run inabout 10 son a 32-bit processcomputer to diag- interrelated in the rugby model and rigid sequencing of
nose a fault as shown in Fig. 4. phases is not intended. The output of the design phase is
The major human resource is needed during the design not a document but a knowledge base that can beexecuted
and debugging of the algorithm. This is easily understood "asis"onthecomputer.Inotherwords,thegoalofthefirst
sincethediagnosisalgorithm cannot bedescribed in a com- phase is to prototype the algorithmas quickly as possible
pact way. As it i s an ill-structured problem, one could not (rapid prototyping). As it i s executable on a computer, both
finish a design document which specified the algorithm user and developer can observe how it works at an early
completely. This results in missing specifications, incorrect stage of the project. If the prototype meets requirements,
code, and frequent modifications as the design becomes the knowledge base is transported to the target machine
(protection ^relay-typo
dz STANDARD EMSHARDWARE
'pro-type local
-id <xl>
^location-at <yl> AI PROGRAY
'objaot <Z>
0 FORTRAN
0 1 WRITTEN I L M R W :
(protectia 'rrhy-type dz
^pro,type
'id
^location,at
local
<x2>
( <y2> <> <yl> }
REAL TYE OS
I
'object <Z>
0
0 1
fault-information
'ar*_rrky nil
'bockup-rahy
dz
local-backup-protection
'spec
'fwlt-at <Z>
0
0 1
1
ubich mems:
"If theback-upr&y <xl> of type dz ot the lo.oation <yl>
that protecta tho eI.nnnt <L> ia opuated d
theback-uprelay <x2> of type dz at tbo location <y2>
to AI PRoGaUls STANDARD EMS H A R D W A R E
that protecta tho elemnt <z> is operated and
<y2> h not q u a l to <yl>.
then tho fault h concluded at tho ohment <I> md
it is clawad by locd back-upprotoction."
1684 PROCEEDINGS OF THE IEEE, VOL. 75, NO. 12, DECEMBER 1987
edge base built into the system. Thus we must be able to [19] C. DeMontravel, “A real time expert system for alarm han-
acquire knowledge efficiently to be able to take the best dling in the futureEDF regional controlcenters,” ClGRE Study
Committee Rep. 39, Sept. 1986.
advantage ofAI.The automatic acquisition such of a knowl- [20] F. Hein, “Expert system using pattern recognition by real time
edge base is discussed in [32] but its state is still primitive signals,” CICRE Study Committee Rep. 39, Sept. 1986.
and knowledge engineers skilled in this area are absolutely [21] K. Matsumoto, T. Sakaguchi, and K. Uemura, “Verification of
necessary. switching operationswith temporallogic,” i n Proc. 8th Power
The knowledgeof how to run a power system is resident System Computation Conf., pp. 1096-1102, Aug. 1984.
[22] C. Fuki and J. Kawakami, “An expert system for fault section
in the engineers and managers of the electric utility com- estimation using information from protectiverelays and cir-
panieswhilethecomputertechnologytobuildanAlsystem cuit breakers,” /€E€ Trans. Power Delivery, vol. PWRD-1, no.
belongs to suppliers. Therefore, close cooperation and joint 4, pp. 83-90, Oct. 1986.
development projectsareabsolutely necessary to reach [23] K. L. Tomsovic et a/., “An expert system as a dispatchers’ aid
forthe isolationof line sectionfaults,” i n Proc. 7986PESTrans-
successful implementations.Finally,peoplehavingthe mission and DistributionConf., Sept. 1986.
multidisciplinary skills of power system engineering, com- [24] S. Mokhtari, J. Singh, and B. F. Wollenberg, “A unit com-
puter science, and cognitive science must be trained for mitment expert system,” i n Proc. 7987 Power lndustry Com-
such tasks. puterApplications Conf. (Montreal, Que., Canada, May 1987).
[25] J. R. Latimer and R. D. Masiello, ”Designof a dispatcher train-
REFERENCES ing simulator system,” in Proc. 7977lEE PowerlndustryCom-
puter Applications Conf., pp. 87-92.
T. Sakaguchi and K. Matsumoto, “Development of a knowl- [26] I. Susumago etal., “Development of a large-scale dispatcher
edge-based system for powersystem restoration,” /€€E Trans. training simulator,” /E€€ Trans. Power App. Syst., vol. PWRS-
PowerApp. Syst., vol. PAS-102, no. 2, pp. 320-329, Feb. 1983. 1, No. 2, pp. 67-75, May 1986.
B. F. Wollenberg, “Feasibility study for an energy manage- [ 2 7 D. Magee et a/., “A large two computer dispatcher training
ment system intelligent alarmprocessor,” i n Proc. 7985Power simulator,” / € € E Trans. PowerApp. Syst., vol. PAS-104, no. 6,
industry ComputerApplications Conf., pp. 249-254, May 1985. pp. 1433-1438, June 1985.
L. R. Larson etal., “The NSP intelligent alarm processor proj- [28] T. E. Dyliacco et a/., “Considerations in developing and uti-
ect,” in Proc.7985 Minnesota Power Systems Conf. (Univ. lizing operator training simulators,” /€€€ Trans. Power App.
Minnesota, Minneapolis, Oct. 8-10, 1985). Syst., vol. PAS-102, no. 11, pp. 3672-3679, Nov. 1983.
A. Doi and K. Uemura, “Knowledge-based operation guid- [29] J.D. Hollan etal., “STEAMER: An interactive inspectable sim-
ance method for voltage and reactive power systems,” pre- ulation-based trainingsystem,” A/ Mag., vol. 5, no. 2, pp. 15-
sented at the IFAC Symp. on Planning and Operation ofElec- 27, Summer 1984.
tric Energy Systems, July 1985. [30] B. Woolf et a/., “Teaching a complex industrial process,” in
C. C. Liu and K. Tomsovic, “An expert system assisting deci- Proc. AAAl Conf. (Philadelphia, PA, Aug. 1986).
sion-making of reactive power/voltage control,”in Proc. 7985 [31] R. D. Masiello eta/., “Review of the EPRl project on operator
Power lndustry Computer Applications Conf., pp. 242-248, training simulator implementation,” presented at the 1984
May 1985. EEI Engineering Computer Forum, Atlanta, GA.
R. Fujiwara etal., “An intelligent load flow engine for power [32] R. S. Michalski,J. G. Carbonel1,andT. M. Mitchell, Ed. Machine
system planning,” in Proc.7985 Power lndustry Computer Learning. Palo Alto, CA: Tioga Publ. Co., 1983.
Applications Conf., pp. 236-241, May 1985.
T. Sakaguchietal., ”Design and developmentof human inter-
face i n engineering workstation for power system planning,”
Bruce F. Wollenberg (Senior Member, IEEE)
presented at the lFAC/lFlP/lFORS/lEA 2nd Int. Conf. on Anal-
was born in Buffalo, NY. He received the
ysis, Design and Evaluation of Man-Machine Systems, Sept. B.S. and M.Eng. degrees from Rensselaer
1985. PolytechnicInstitute, Troy, NY, andthe
Y. Kohno etal., “An intelligent supportsystem for powersys-
Ph.D. degree from The University of Penn-
tem planning,” in Proc. 7985 lnt. Symp. on Circuits and Sys-
sylvania, Philadelphia.
tems. Also i n /€E€ Trans. Power Syst., vol. PWRS-1, no. 2, pp.
He has held positionsatLeeds and North-
67-75, May 1986.
rop Co., North Wales, PA, and Power Tech-
R. L. Moore etal. ”A real time expert system for process con-
nologies Inc., Schenectady, NY. He is pres-
trol,” i n Proc. ?st /€€€ Conf. on Artificial Intelligence Appli-
entlywith theEnergy ManagementSystems
cations, IEEE publ. CH 2107-1/84, pp. 519-576.
Division of Control Data Corporation, Ply-
T. E. Murphy, ”Developing expert system applications for
moth, MN. He has served as an adjunct faculty member in Elec- the
process control,” in Proc. 7985 /SA Conf. (Philadelphia, PA).
tric Power Engineering Department at Rensselaer and in theElec-
P.A.Sachs et a/., “Escort-An expert system for complex
trical Engineering Departmentat the University of Minnesota. His
operations i n real time,” Expert Syst., vol. 3, no. 1, Jan. 1986.
research interests include power system operations, large-scale
D. G. Cain, “BWR shutdown analyzer using artificial intel-
optimization of power system problems, and application of arti-
ligence
techniques,” Rep. NR-4139-SR, Electric Power
ficial intelligence techniques t o energy management systems.
Research Inst., July 1985.
C. L. Forgy, “OPS5 user’s manual,” Tech. Rep. CMU-CS-81-
135, Dep. Comput. Sci., Carnegie-Mellon Univ., Pittsburgh,
PA, July 1981.
B. D. Clayton, “ART programming tutorial,” InferenceCorp., Toshiaki Sakaguchi (Member, IEEE) received
1984. the B.S. and M S . degrees i n 1969 and 1971,
R. A. Kowalski, Logic for Problem Solving. Amsterdam, The respectively, and the Ph.D. degree i n elec-
Netherlands: North-Holland/Elsevier, 1979. trical engineering i n 1981, all from Kyoto
R. P. Shulte et a/., ”Artificial intelligence solutions t o power University, Kyoto, Japan.
system operating problems,” paper 86 SM 3354, presented He joined Mitsubishi Electric Corpora-
at the 1986 IEEE Power Engineering Society Summer Meet., tionin 1971, and has beenworkingin
Mexico City, Mexico. research and development on power sys-
S. N. Talukdar and E. Cardozo, “Artificial intelligence tech- tems technology at the Central Research
niques for power system operations,” Rep. EL-4323, Electric Laboratory. He i s currently the Research
Power Research Inst., Jan. 1986. Managerthere. Hiscurrent interest is i n the
H. Kaninsono, “Alarm handling by control and load dispatch- analysis of human heuristics and their use for management and
ing centers in Japan,” ClGRE Study CommitteeRep. 39, Sept. control of large-scale systems.
1986. Dr. Sakaguchi is a member of the IEE of Japan.