Engineering, 2013, 5, 191-196
doi:10.4236/eng.2013.51b035 Published Online January 2013 (
Copyright © 2013 SciRes. ENG
Approach of Training Working Staff of Power System
Operation Mode Based on State Evaluation
Qian Feng, Fan Youping, Liu Chen ,Ai Yunfei
Power Dispatching and Communication Center Of Guangdong Province, Guangzhou,China
Received 2013
An approach of training working staff of power system operation mode based on state evaluation is proposed. In terms
of knowledge features of power system operation mode, we studied a training method based on evaluation of learning
state. This training method makes individual learning for different individual condition to give them ability to grasp
learning points quickly, evaluate real-time learning effect, update learning style in time and summarize problems after
one learning stage, so that learners can master professional knowledge in constant summaries and feedbacks. Obvious
effects can be obtained on discontinuous learning time that trainees can master basic theories associated with their
working and operations adapted to practical work quickly.
Keywords: Power System Operation Mode; P rofessional Traini ng; Sta te E valuatio n
1. Introduction
Operation mode department assumes important task in
power system, constantly requiring a large number of
calculation and analysis to ensure security and stability
of power system operation. At present, practical training
effect of operation mode is not that ideal. Such training
just crammed the knowledge to learners that acceptance
and mastery couldn’t be reflected in time. Since base and
professional background of working staff vary, the train-
ing emphasis is different for each trainee.
As the notion of scientific training becomes popular,
staffs, especia lly knowledge sta ffs, pa y more attention to
get scientific and efficient knowledge training provided
by the entity. This training state evaluation method ex-
actly provides a good insight into improving training
effect and reconstructing more scientific and more prac-
tical training system.
2. Basic Model of Scientific Training
2.1. Basic Model of Training System
In the whole trainin g system, training e valuation is a very
critical component. Without it, the system couldn’t be
intact. An intact training system model is shown in Fig-
ure 1.
A scientific training system analyzes organization,
work and individual to determine training demand. Then
it identifies objectives, by which it can determine objects,
contents, time and ways. After that, it draws out training
program, which is in fact the specification and operation
of training obj ectives. Ne xt is to carr y out training activi-
ties and final to evaluate.
2.2. Model of Training Cyclic Process
Training process should be a systematically cyclic
process. In this cyclic system, both evaluation and train-
ing demand analysis are important parts, belonging to
independent core part. When there was no training eval-
uation, cycle would not occur, so that training is directly
from demand analysis to implementation. Figure 2 shows
a simplified training cyclic system model.
Demand determine
Training evaluation
Situation analysis:
Organization Work
Figure 1 . b a s i c model of training sy stem.
Copyright © 2013 SciRes. ENG
Considering training evaluation as part of the cycle
will contribute to the entity’s feeding back practical op-
eration results to decision-making layer by monitoring
and correcting of training process. After the feedback, we
compare results with predetermined objectives to find out
deviations and then adjust and correct practical work to
make i t go o n whe el s on o r igi na l p la n o r a d j uste d p lan by
a series of judgments.
2.3. Scientific Training Evaluation Method
D.Kirkpartrick divided training effect into four progres-
sive levels that are reaction, learning, action and result in
terms of evaluation depth and difficulty, and provided
answers to four questions (see T able 1).
2.4. Particularity and Demand of Training
Working Staff of Power System Operati on
Training simulators of po wer system at home and abroad
have been developed since 1970s. it started focusing on
generation units of thermal power plants, then extending
to si mulation tr aini ng on powe r network a nd substa tions.
Dispatcher Training Simulation (DTS) is widely used in
all national dispatching centers as a powerful tool. Subs-
tation Training Simulation System (STS) needs to not
only build realistic working environment and simulate
potential contingencies of substation as well as perform
nor mal op era tions and c onti nge ncy ha ndli ng t raini ng, b ut
also be able to make full use of computer technology to
implement training fr om various aspects.
Figure 2 . S implified training cyclic system model.
Table 1. Four-level evaluation met hod of D.K irkpartrick.
rank questions
1.reaction Which aspect s of traini ng progr am can meet train ees’
2.learning What did traine es gain fr o m thi s p r o gram?
3.action Has trainees ’ behavior chan g ed through training?
4. r esult
Did beh avior change have p ositive impac t on organi-
Work of power system operation mode involves more
theoretic application due to its particularity. Therefore,
its training is different with conventional training. The
working characteristics are mainly as following:
(1) Stro ng theore tic a l
Work of power system operation mode involves more
theoretic application, such as power network structure,
power flow calculation, security and stability analysis
and judgment etc. For working staff of operation mode,
mastering basic theoretic knowledge of power system
well is the base of effective work.
(2) Many involved software operating
Power system operation mode has to analyze network
structure of this province or this gird constantly, while
complexity of which determines that working staff must
skillfully use power syst em c omputation software suc h as
BPA. They can make right judges quickly and submit
analysis report at first time only if they mastered com-
mon computation software.
(3) Busy work
Operation mode department has to analyze future grid
state in advance, and make corresponding arrangements
like expansion or maintenance, so work of this depart-
ment is ahead of grid operation. When there are impor-
tant activities, festivals or critical maintenance and ex-
pansion, operation mode department needs to submit
anal ysis re p or t i n ad va nc e, whi ch mea ns t ha t wor k o f t hi s
department is busy.
Thus, training for operation mode needs a systematic
and individual method. In the new approach, professional
points can pass to trainees visually and accurately for
easy understanding; trainees can arrange time at their
own wills and save learning progress to provide feed-
backs for state evaluation so that improve learning effi-
3. Training Method for Working Staff of
Power System Operation Mode Based on
State Evaluation
3.1. Overall Framework of Training
Combined with scientific training system model, this
paper proposed an approach of training working staff of
power system operation mode based on state evaluation
in terms of the working characteristics. Overall training
process is shown in Figure 3. This method consists of 4
components: state analysis of training effect, that is, di-
viding process into three stages and setting learning state
variables for each stage to analyze and test; e valuation o f
specific learning state, which represents by combining
training software evaluation, trainee’s self-eval uat io n a nd
trainer’s evaluation to evaluate each stage objectively;
updating of training state, implemented by training soft-
ware’s resulting state historical database; Cumulative
Copyright © 2013 SciRes. ENG
assessment o f training state, i n other words, to divide the
training into 4 stages to per form differe nt level of evalua-
tion, respectively, and finally make total assessment.
3.2. Evaluation of Specific Learning State
Evaluation of learning state is divided into three stages:
state examination before learning, state evaluation during
learning, state updating after learning. Distribution is as
Figure 4 shows.
1) Sta te examination befor e training
This stage includes two parts that are readiness exams
of basic knowledge and practical skill. Since the basis
varies, examining readiness of trainees can give a rough
idea of their knowledge and skill base to make training
Demand determine
Training target
Training evaluation
Evaluation before training
Requirement of the
professional knowledge
Learning foundation
Historical databaseSave
Figure 3 . Process of overal l training method.
Learning state of
before training
State evaluation
during trainingState updating
after training
knowledge Practical
Response state
of trainees
Learning effect
Knowledge and
ability state
Figure 4. Distr ibuti on of learni ng st a te.
Readiness exam of basic knowledge is mainly about
theoretic knowledge. Before training, we can know about
knowledge readiness of trainees by a quick electronic test.
Test content is knowledge points of courses to be taken,
and test questions are harder as the question number in-
creases. In the end, trainees’ readiness is graded accord-
ing to their answers. Mastering level of knowledge de-
cides the amount of energy they should pay to courses to
be trained. If a trainee has already gotten corresponding
professional knowledge base, he can jump this part.
Readiness exam of practical skill is more complex.
Because of the little ti me available for exam, this exam is
mainly about self-test. Before training one skill, such as
use of BPA and setting sheet formation of spare power
automatic s witching device, trainees get fa miliar with the
soft war e on t hemselves a nd se lf -eval uate their familia rity
and proficiency. Self-test is done with software function
blocks to let trainees judge their behavior. After test, the
system will ask trainees to self-grade in terms of given
levels, thus obtain the rank of skill readiness. System
then provides purposeful training for each trainee by the
self-test results. If someone has a good skill base and
reaches a good mastering level, he can jump this part.
2) Sta te evalua tion during training
Response state and learning effect state during training.
In the training, trainees’ response reflects their learning
effect. If they respond quickly and actively, they have a
good learning state and are easy to accept knowledge;
while if t he y re spond slo wl y, the y ha ve n’t full y ma ster ed
the knowledge, so they need to continue consolidating
what they have learned. In this stage, after single know-
ledge point is taught, training system will pop up choice
questions. Through trainees’ response condition and
answering time, their response state can be provided to
determine learning effect. If response to one point is slow
and accuracy is not high, it demonstrates that the trainee
has some difficulty in receiving this point and system
will repeat it again more slowly. To each test, system
will record corresponding state so that can evaluate
learning state after training.
3) Sta te updating after training
Overall test of entire knowledge module is needed af-
ter training, which is a test with time limit. Simple basic
knowledge has a short time limit, requiring trainees to
give answers quickly at first time; difficult essay ques-
tions and calculatio n questions also ha ve time li mits cor-
responding to difficulty level, requiring trainees can
master the points well and apply them and skills fast.
Test result after training is measured with two variables
that are total score and answering time. Although one
may get a high score, if he gave answers near or on the
time limit, he is demonstrated to just master for tempo-
rary and system still suggests the trainee to consolidate
after trainin g.
Copyright © 2013 SciRes. ENG
After the training, the summary can be done as
four-stage evaluation of D.Kirkpartrick (Table2). Incor-
porating specific training content, we can evaluate
whether the training has prominent effect on improving
trainees’ knowledge and sk ill state.
3.3. State Updating
Evaluation process of learning state is shown in Figure 5.
In the training, state variables of each stage for trainees
reflect specific condition of entire training and are rec-
orded and preserved for future training.
Table 2. Four-stage state evaluation after training.
rank questions
1.reaction Are trainees satisfied with knowledge points in training
2.learning Have tra inees ma stered kn owledge th eory or op erati on
ski lls f rom training program?
3.action Does behavior in working change through training?
4.result Does beh avi or chan ge cont rib ute t o suc cess ful workin g
of operation mode depa rtmen t?
Gathering training state
Evaluation of teachers
Self-evaluation of
Problem in the training
State examination before
Targeted learning
Test evaluation and state
updating The 3 stages of
state evaluation
Figure 5. Evaluation proce ss of learning state.
System files states of each training class which will
accumulate to form training state historical data base. To
every stage in training, we compare the state variables of
the same stage (state before, during, after training) in
dif fere nt kno wled ge t ra ini ng t o get c ha nge cur ve. W he the r
a trainee has already mastered knowledge is judged from
the curve. Learning state is good with uptrend, that is,
someone has well digested what we taught, while bad
with violation, meaning that problem is to be found and
state should be adjusted.
3.4. State Accumulative Evaluation
As figure 6 shows, training consists of three knowledge
subsystems: basic training system, professional skills
training system, as well as working process and man-
age ment tr aining system.
Basic training introduces criteria and basic knowledge
associated with operation mode, stage evaluation of
which is in for m o f electronic questions.
Skill training involves the use of BPA and so on.
Working process and management is about operation
mode reports of each period, analysis of power plant’s
switching into system, electricity guarantee scheme,
generation source management and operation manage-
ment of stability control equipment.
The second and third subsystems have to set tests for
professional skills in addition to electronic questions.
Professional skills are tested from multi-aspects. Training
software simulates specific working scene for trainees to
judge and respond to test their ability of knowledge ap-
plication and quick response.
Basic training
Professional skills training
Working process and management training
Stage test
Stage test
Stage test
Overall assessment
Figure 6. Stage accumulative eval uation process of training
Copyright © 2013 SciRes. ENG
Overall assessment is done after all training of the
three knowledge subsystems. Basic theoretic knowledge
is permeated in comprehensive questions and applica-
tions to test trainees’ ability of applying knowledge. Af-
ter all training stages, trainees can go back to previous
training blocks to review knowledge and consolidate
what they have learned by constant feedbacks.
4. Fuzzy Comprehensive Assessment
Fuzzy comprehensive assessment is to quantize practical
fuzzy variables using assemble and fuzzy math. This
assessment can evaluate trainees during training and
monitor trainers. Consider skill training of BPA calcula-
tion software. It illustrates the process of fuzzy compre-
hensive assessment during traini ng.
In terms of features of BPA skill traini ng, tasks suc h as
data input program, power flow calculation and short
circuit current calculation are designed as corresponding
knowledge p oints.
To every point, there are three test forms.
completion ; simulate operation test on graphical in-
terface; error correction. Through three tests above,
we can exam mastering condition of trainees to provide
basis for next dec ision-making.
Specific procedure of comprehensive assessment is:
(1) During test, if a trainee answers one question cor-
rectly, the corresponding cognitive ability or abilities
would be 1; if wrongly, be 0. Which terms the question
effects is given by experts. After the trainee finishes all
questions associated with one point, he can obtain a test
record table as table 3 shows.
Table 3. knowledge t es t record table.
Reflections of each cognitive ability in training
number memorization comprehension application analyzation integration evaluation
1 1 0 1 1 1 0
2 1 1 1 0 1 1
3 1 0 0 1 1 1
… … … … … … ……
n n n n n n n
(2) After test, representation of each terms obtained
from trainee to some type of test can be provided.
[0,1],0 6,
z iz∈ ≤≤
indicates value of one
cognitive ability.
where, 0≤j≤n (n is the number of test questions).
is the times of corre ctly answering ith cognitive ab ility in
this test. Cognitive ability includes six ter ms of memori-
zation, comprehension, application, analyzation, integra-
tion, evaluation.
(3) Test pattern of each knowledge element involves
the three forms previously described. So, evaluation ma-
trix of mastering degree to each test can be constituted
after test.
11 1216
21 2226
31 3236
zz z
Az zz
zz z
Where, the first row represents the six evaluation val-
ues of mastering degree in completion. The second row
represents that in simulated operation. The third row
represents that in error corr ection.
(4) Weights of three for ms are determined by
T1 is the weight of completion; T2 is the weight of
simulated ope ration; T 3 is the weight of erro r corre c tion.
/( )
Ttttt=++ (5)
Where, t1, t2, t3 are the average answering evaluation
value of one test form after several tests, respectively.
They are pre-set by experts at the beginning use of sys-
tem and can be adjusted correspondingly with training
effect after some time of training test.
(5) Calculate final mastering level after test. Evalua-
tion results are:
11 1216
1 23212226
31 3236
(, , )
zz z
zz z
= =
 
(6) Calculate trainee’s mastering condition M of this
knowledge p oint:
Where, Qi is the weight of mastering ability of one
knowledge p oint.
5. Conclusions
Using scientific training system model, we developed an
Copyright © 2013 SciRes. ENG
approach of training working staff of power system op-
eration mode based on state evaluation, effectively over-
coming problems of busy work as well as non-systematic
and concentrated training. Fuzzy comprehensive evalua-
tion is applied in state evaluation, which dealt with state
data of test after training and reflects training state in
time to i mprove learning e fficiency.
[1] Donald Kirpatrick. Techniques for Evaluation Training
Programs [M], Training and Development v50 p54-9
[2] Laurie J.Bassi, George Benson, and Scott Cheney. The
Top Ten Trends [M],Training and Develop-
ment(Ale xandria,Va.) v50 p38 -54 N’96.
[3] ZHANG Yue. Fuzzy mathematics method and its appli-
cation [M]. China Coal Industry Publishing House, 1996.
[4] TIAN Ling. Research on Fuzzy Comprehensive Evalua-
tion of Electric Energy Purchasing Scheme Based on
Analytic Hierarchy Process [J]. Power System Technolo-
gy, 2005,29(7):23-27.
[5] QI Hong,YUAN Wei. Research on Distance Learner's
Learning Characters and Strategies [J]. Distance Educa-
tion in China, 2005, (17): 34-41.
[6] LI Haiyan. Distance Learners' Behaviors at Different
Stages of the Learni ng Process [ J]. Di stance Educati on in
China, 2009, (3) : 4 4-47.
[7] GAO Bo. The Research of Application of Meta-cognition
and Fuzzy Theory in Electric Power System Simulating
Training [D]. North China Electric Power University,
[8] CAO Ronghui, DONG Hongwei. Remote Education
Training System for Power System Based on IP [J].
Northeast Electric Power Technology, 2006,27(12): 31-3