Research on Post Evaluation Method for Power Grid Technology

Projects Based on AHP-CRITIC-TOPSIS

Jianjun Wang

*

, Qidi Zhao, Cunbing Li and Jiayin Pan

School of Economics and Management, North China Electric Power University, Beijing, China

Keywords: Power Grid Technology Project, Post-Evaluation, AHP, CRITIC, TOPSIS.

Abstract: The power system serves as a crucial foundational measure for facilitating national economic development

and ensuring convenient energy access for the people. In order to promote China to achieve carbon peak and

carbon neutral vision, China is actively promoting the development of a new low-carbon power system. To

meet the urgent demand for precise investments in the power grid within the context of carbon peak and

carbon neutrality, there has been a significant increase in the number of power grid technology projects. The

scientifically objective post-evaluation methods for power grid technology projects are of utmost importance

in effectively guiding the innovative development of these projects. This paper focuses on the post-evaluation

of power grid technology projects, aiming to provide valuable research support in this field. Based on the

relevant concepts and objectives of post-evaluation for power grid technology projects, eight indicators are

selected from the four aspects of organizational management, technical level, achievement application and

influence to construct an indicator system for the post-evaluation of power grid technology projects.

According to the value of indicators, an evaluation model of post evaluation of power grid technology projects

is constructed by using AHP and CRITIC subjective and objective weight method. Additionally, we validate

the effectiveness of this approach using a case study of typical technologies in the new power grid. The

research of this paper can provide quantitative reference for promoting the innovation and development of

power grid technology projects.

1 INTRODUCTION

In the face of the global challenge posed by climate

change, the international community has reached a

broad consensus and taken concerted action.

Currently, close to two-thirds of nations have

explicitly set carbon peaking and carbon neutrality

goals, signifying the emergence of a global trend

towards low-carbon transformation. In order to

advance sustainable development, China has set forth

the objectives of achieving carbon peaking by 2030

and carbon neutrality by 2060. As an industry with

high carbon emissions, the electric power system is in

urgent need of accurate investment in power grid

under the background of carbon peaking and carbon

neutrality. At present, the number of power grid

technology projects has increased significantly, so a

scientific and technological post-evaluation method

for assessing the scale and speed of power grid

development is essential, structure and safety,

efficiency and benefit under the new power system

characterized by the integration of low-carbon green

development, operational efficiency and economic

benefit.

The power grid innovation technology has a

significant and far-reaching impact on the power grid

enterprises, and will even influence the development

of energy and economy. So, it is significant to develop

the study on the impact assessment of power network

innovation technology to guide power network’s

development and establish the innovation mechanism.

Many developed countries and top scientific and

technological enterprises have carried out relevant

research.

The development of smart power grids is leading

to the transformation of the power system into a new

intelligent system. Therefore, many scientific and

technological projects are mainly evaluated and

studied based on intelligence. For example, IBM has

identified five stages for the construction of a smart

power grid, which are used to assess the maturity of

innovative technologies in developing a smart power

grid. These stages focus on improving the reliability,

efficiency, acceptance of new energy, and interaction

ability of the smart power grid. The constructed

262

Wang, J., Zhao, Q., Li, C. and Pan, J.

Research on Post Evaluation Method for Power Grid Technology Projects Based on AHP-CRITIC-TOPSIS.

DOI: 10.5220/0012280500003807

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Seminar on Artiﬁcial Intelligence, Networking and Information Technology (ANIT 2023), pages 262-270

ISBN: 978-989-758-677-4

evaluation system selects 8 items and about 200

indicators to find differences through comparison and

direct the development of power grid development.

The U.S. Department of Energy evaluates the impact

of innovative technologies on smart grids from six

perspectives: user involvement, new products’

introduction, power grid operation efficiency, quality

of power service, energy storage devices, and power

grid disaster prevention capability. On the basis of this

evaluation indicator system, the American Institute of

Electrical Science and Technology further expanded

the evaluation index system of a specific grid

construction project, and refined the index system of

the Department of Energy, to evaluate the impact of

innovative technology on the benefits and

development of construction projects. For the purpose

of building smart grids, namely, to realize low-carbon

economic development by increasing the connectivity

of renewable energy sources such as wind power and

promoting the use of technologies for distributed

power generation and demand-side management,

Europe has established an impact evaluation system of

innovative technologies on smart grids and made use

of KPI theory. A total of 21 KPI indicators were

extracted from the perspectives of sustainable

development, power transmission, power grid access

standards, safe and high-quality power supply, power

grid operation efficiency quality, networking ability,

coordinated planning and development, cost

efficiency, innovation ability, etc., to evaluate the

impact of smart grid technology, equipment,

interaction, and revenue ability (Amin D, 2021).

Researchers in Brazil considered 13 technical and

economic criteria to investigate a multi-criteria

approach and developing an expert system-based

computational model to assess the effectiveness of

distribution network operators in Brazil (Ghizoni C R

T, 2022). Based on the TSFPMSM operator, a

MAGDM algorithm was developed for the evaluation

of Pakistan's smart grid network. Based on the

observation of the response to changes in sensitive

parameters, the proposed numerical examples are

subjected to sensitivity analysis, and a comprehensive

comparative study is conducted (Areeba N, 2022).

In addition, there are many related studies in the

field of power system evaluation. Xiufan M

established a comprehensive set of evaluation

indicators across five dimensions: reliable operation,

economic performance, efficient interaction,

technological intelligence, and green emissions

reduction. A comprehensive evaluation model for a

5G+ smart distribution network was proposed based

on cloud modeling, which incorporates the principle

of minimizing variance and accounts for the

uncertainty of information pertaining to distribution

network nodes and equipment statuses0. Long C W

introduced a new analytical model and relationship

assessment method that takes into account grid

evolution, integrating both rapid dynamics and slow

evolution. This model encompasses load increases,

upgrades, and construction of equipment such as

power plants, transformers, and transmission lines,

simulating the development of the power grid by

modeling 0. However, with the release of the "carbon

peaking and carbon neutrality" action plan and the

continued drive for energy transformation,

accelerating the establishment of a clean, low-carbon,

secure, and efficient energy system while continually

advancing carbon reduction has become the next

crucial focus. To achieve this goal, the power grid

needs to undertake a significant number of technology

projects and research as support. A multitude of

technology projects not only facilitates the rapid

advancement of low-carbon power generation

technologies but also develop new energy projects

according to local conditions based on the advantages

of each province. In the face of this vast array of

technology projects, Faced with a far larger number of

technology projects, it is essential for the power grid

to conduct systematic screening and evaluation.

Therefore, establishing a post-evaluation system for

technology projects is necessary to identify the most

valuable initiatives to pursue, ensuring that the

research outcomes from these technology projects can

assist the power grid in addressing technical

challenges and achieving resource allocation and

optimization during the transition. This will support

the power grid in completing energy transformation

and promoting the low-carbon transformation of the

power system. It can provide information support and

reference for project investment decision and process

management.

2 CONSTRUCTION OF POST-

EVALUATION CONCEPT AND

EVALUATION INDEX SYSTEM

FOR SCIENCE AND

TECHNOLOGY PROJECTS

2.1 The Concept of Post-Project

Evaluation

Post-evaluation of science and technology projects

refers to the activities of comprehensively analyzing

and evaluating the implementation process, benefits

Research on Post Evaluation Method for Power Grid Technology Projects Based on AHP-CRITIC-TOPSIS

263

and internal and external influences of the projects by

using scientific and systematic evaluation methods

after the projects are completed or put into operation.

Through the analysis of the actual completion and

operation of the project, it can compare whether the

project has reached the predetermined target when the

project was set up in terms of output, benefit and other

indicators, and make a scientific evaluation by

comparing the completed target and the predetermined

target; Analyze the decision-making process and

implementation process of the project, find the

existing problems, summarize the experience and

lessons, and provide feedback. According to the

definition of post-evaluation, it can be interpreted

from the following three aspects: From a purpose

standpoint, post-evaluation of technology projects

serves as the primary approach by which project

management departments manage and assess

technology projects, and is also an important means of

science and technology project management. Its main

task is to evaluate the benefits of science and

technology projects in an all-round way, and to feed

back the evaluation results to the science and

technology management department, so as to provide

a basis for the modification of science and technology

project management mode and policy. Secondly, from

the perspective of stages, the whole process

management of science and technology projects is

composed of post-evaluation, project selection

demonstration, project evaluation, mid-term

inspection, acceptance appraisal and other stages.

Although post-evaluation constitutes the final phase of

a technology project, its significance is paramount

within the entire lifecycle management of such

projects0. Only after evaluation can science and

technology projects accurately reflect the long-term

impact of results. Finally, from the perspective of

object, post-project evaluation of science and

technology can evaluate a single project or multiple

projects under a certain type of special plan.

2.2 Power Grid Science and Technology

Project Post-Evaluation Content

According to the characteristics of the project, the

content of post-evaluation of power grid science and

technology projects is evaluated from the aspects of

the completion of the project objectives, the project

implementation and management, the comprehensive

benefit of the project results, the comprehensive

influence of the project, the technical level and

application of the project results.

(1) Achievement of project objectives

By comparing some economic and technical

indicators actually produced by the project with the

goals determined during the project decision-making,

the project can be checked whether it has reached the

expected goals or the degree to which it has reached

the goals, and the deviation can be analyzed to judge

the success of the project. Additionally, it is necessary

to analyze and assess the effectiveness, reasonability,

and feasibility of the initial project decision

objectives.

(2) Project implementation and management

status

By analyzing whether the project decision is

scientific and feasible, whether the resource

investment can be further optimized, whether the

project scheduling is reasonable and other aspects, the

fine degree of management in the process of project

implementation is evaluated, and the deficiencies in

management organization are found and solutions are

proposed.

(3) Comprehensive benefits of project results

From the perspective of economic benefit,

according to the actual input and income data of each

year during the post-evaluation, the economic benefit

is evaluated; from the perspective of social benefits,

whether to support the implementation and theory of

major national policies, whether to lead or open up

new technical fields, to help the company's high-

quality development; and consider the comparison

with the pre-project assessment, identify the reasons

for the significant changes, and summarize the

experience and lessons.

(4) Comprehensive project influence

After the project is completed and put into

operation, an assessment should be conducted to

evaluate its actual impact on the local economy,

society, and environment. Based on this assessment,

the project's decision objectives should be determined,

including evaluations of its economic impact, social

impact, and environmental impact.

(5) Technical level and application of project

achievements

Judge whether the technical level of the project

results is advanced enough, whether they can be

applied according to the current characteristics and

have enough applicability; Whether the technology of

the project results is mature enough, whether it reaches

the expected goal, and whether the application method

is clear.

Construction of post-evaluation index system for

power grid science and technology projects

The post-evaluation of power grid science and

technology projects generally needs to achieve four

ANIT 2023 - The International Seminar on Artiﬁcial Intelligence, Networking and Information Technology

264

objectives: 1) Better understanding of the influence of

competition policy decisions; 2) Improve the

transparency and accountability of competition policy

decisions; 3) Promoting competition and competition

policy; 4) Improve future decision-making practices.

Therefore, we understand the objective of post-project

evaluation as follows: through a comparative analysis

between the project's anticipated objectives and actual

outcomes, comprehensively summarizing the project

implementation process, drawing lessons from it, with

the aim of improving project decision-making, design,

execution, and management, and ultimately achieving

the project's anticipated objectives.

Table 1. Evaluation index model of basic prospective

scientific research projects.

Serial

number

First-order index Secondary index

1

Organization

management

1

A

Project completion

schedule

1

B

2

Project acceptance status

2

B

3

Technical level

2

A

Technological maturity

3

B

4

Project approval accuracy

4

B

5

Application of results

3

A

Intellectual property

rights

5

B

6

Technical support

6

B

7

Influence

4

A

Project extension

7

B

8

Personnel training

situation

8

B

From the above definition and the objective of

post-evaluation, the post-evaluation indicators of

power grid science and technology projects are

selected from four dimensions: organizational

management, technical level, achievement application

and influence, and a number of supporting indicators

are set under each dimension (a total of 8 supporting

indicators) to improve the degree of refinement and

comprehensiveness of post-evaluation. The evaluation

index model of basic prospective scientific research

projects is shown in Table 1.

3 CONSTRUCTION OF GRID

SCIENCE AND TECHNOLOGY

PROJECT POST EVALUATION

MODEL BASED ON AHP AND

CRITIC

3.1 Index Weight Calculation Method

The post evaluation index system of different types of

scientific research projects established in this paper

not only includes the subjective evaluation method

based on AHP, but also includes the objective

evaluation method based on CRITIC method. The

evaluation index system for various types of scientific

research projects is a complex and dynamic system

containing fuzziness and accuracy at the same time,

with a variety of factors, certainty and uncertainty. If

only one evaluation index is considered, it is difficult

to obtain comprehensive evaluation results. It is a

necessary feature of weight determination method to

reflect the fuzziness and correlation among evaluation

indexes. Therefore, in the comprehensive evaluation,

it is more appropriate to adopt the subjective and

objective weight combination calculation method

combining AHP method and CRITIC method. This

method quantifies the weight of various evaluation

indicators, making the comprehensive evaluation

results have obvious rationality.

(1) Subjective evaluation method based on AHP

The specific operation steps of AHP are shown as

follows0:

1) Establish the hierarchical structure model

AHP can simplify complex problems by layering,

and divide factors into different levels according to the

interrelation and dominance of various factors.

2) Construct a comparative judgment matrix

Judgment matrix is a matrix constructed by

decision makers in judging the mutual importance of

elements of each layer in the index system according

to certain methods. Judgment matrix

B

is as follows:

11 12 1

21 22 2

12

n

mm mn

bb b

B

bb b

















(1)

ij

b

indicates the importance of elements p

i

and

j

p

relative to the criterion

k

C of the upper layer. In this

study, the value of

ij

b

comes from researchers with

senior research experience in related fields. Therefore,

the preliminary judgment matrix

B

is consistent, that

is, it meets the following conditions:

Research on Post Evaluation Method for Power Grid Technology Projects Based on AHP-CRITIC-TOPSIS

265

1ij ijbb

,

1

ji

ij

b



,

(, , 1,2 )

ik

ij

jk

b

bijk n

b

，，

(2)

The evaluation indicator system established in this

paper includes many qualitative indicators, and the

evaluation of the importance of different indicators

comes from the subjective judgment of researchers in

this field. In order to make the qualitative data easier

to be quantified, this paper adopts the 9-level scale

method to determine the importance of each indicator

ij

X

, as shown in Table 2.

Table 2. Evaluation criteria of the grade scale method.

Assignment of

ij

b

How important

i

x

is compared to

j

x

1

i

x

is of equal importance to

j

x

3

i

x

is slightly more important than

j

x

5

i

x

is more important than

j

x

7

i

x

is very important than

j

x

9

i

x

is extremely important over

j

x

2，4，6，8

The corresponding transition scale between the

preceding and the following two stages

Reciprocal

Scale of importance of

i

x

over

j

x

3) Calculate the weight of each layer

①Multiply each row of elements of the judgment

matrix:

1

,( 1,2, , )

n

iij

j

mbi n



  

(3)

②Calculate the NTH root of m

i

to get the feature

vector

i

w :

,( 1,2, , )

n

ii

wmi n

(4)

③The vector

12

),,,

n

Www w（ is normalized:

1

/(1,2,,)

n

ii i

i

Ww wi n









(5)

12

),,,

T

n

Www w（

is the approximate solution

to the eigenvector.

4) Consistency check

①Calculate the maximum characteristic root

max



:

max

1

n

i

BW

nw







(6)

Where,

B is the judgment matrix and W is the

weight vector.

②Calculate the matrix consistency index

CI :

1

max-n

CI

n







(7)

③Calculate the consistency ratio

CR :

CI

CR

RI

 (8)

Where,

RI

is the average randomness index,

whose value is selected from the table given by

Thomas (1986) (TABLE 3). The judgment criteria are

as follows: when

CR is less than or equal to 0.1, the

matrix has consistency, indicating that the consistency

test is passed, and then the next operation can be

carried out, that is, to sort the weight. When the value

is greater than 0.1, it means that the judgment matrix

does not have good consistency and cannot pass the

consistency test. At this point, the matrix needs to be

modified and the weight value of the judgment matrix

reevaluated until the consistency criteria is passed.

Table 3. Values of RI indicators.

n 1 2 3 4 5 6 7 8 9 10

RI 0 0 0.57 0.87 1.12 1.26 1.36 1.41 1.46 1.49

5) Calculate the final weight set

The premise of this step is that the judgment matrix

has passed the consistency criteria. At this time, the

weight of each evaluation index in the middle layer for

different types of scientific research projects can be

calculated. The formula is:

iijWWW



 (9)

Where,

1,234i



,, and

1, 2, 3, 4, 5j 

.

(2) An objective evaluation method based on

CRITIC method

Under normal circumstances, objective weighting

method is based on sample data. Coefficient of

variation, standard difference and other values are

used to represent the information content of each

index, and index weights are allocated according to

this standard. However, the CRITIC objective weight

method not only takes into account the amount of

information carried by each indicator, but also takes

into account the problem of information duplication

caused by the correlation between indicators. The

specific calculation steps are as follows:

Under normal circumstances, objective weighting

method is based on sample data. Coefficient of

variation, standard difference and other values are

used to represent the information content of each

index, and index weights are allocated according to

this standard. However, the CRITIC objective weight

method not only takes into account the amount of

ANIT 2023 - The International Seminar on Artiﬁcial Intelligence, Networking and Information Technology

266

information carried by each indicator, but also takes

into account the problem of information duplication

caused by the correlation between indicators. The

following are the specific steps for performing the

calculation:

1) Dimensionless processing is carried out for

each index

The process of dimensionless processing of sub-

indexes can neutralize the influence of varying

dimensions on the assessment outcomes. Under

normal circumstances, forward or reverse treatment is

selected for the CRITIC method, and standardization

processing is not recommended, because standardized

treatment will cause all the standard deviations to

become 1. In this way, all indicators will have

completely consistent standard deviations, and the

volatility indicator is meaningless.

Forward or reverse processing:

When the value of the used index is larger, the

better (positive index):

min

max min

j

ij













(10)

When the value of the index used is as small as

possible (inverse index):

max

max min

j

ij









(11)

2) Index variability was calculated

The

jth index is expressed as

j



in the form of

standard deviation, and the calculation formula is as

follows:



2

1

n

ji

i

n











(12)

Where,

i



is the ith value of index

j

,



is the

arithmetic mean of

i

x

,

n

is the total number of

i

x

.

The standard deviation is employed to quantify the

internal dispersion of numerical values among various

indicators, thus reflecting the disparities in values

within each indicator. A larger standard deviation

indicates greater numerical variations among the

indicators, signifying a wealthier information content

within these indicators and a higher evaluative

significance. Consequently, indicators with larger

standard deviations should be assigned a greater

weight in the evaluation process.

3) Calculate the index conflict

Expressed by correlation coefficient, the

quantization formula of the conflict between the

jth

index and other indicators is:

1

(1 )

n

j

ij

i

Rr







(13)

Where,

ij

r

evaluates the correlation coefficient

between index

i and

j

. Correlation coefficients are

used to measure the degree of interrelation between

indicators. The stronger the correlation, the less

conflict exists between this indicator and others, and

the more redundant information is present. This

redundancy leads to repetition in the evaluation

process, consequently weakening the evaluation

strength of the indicator. Therefore, when balancing

the importance of evaluation indicators, it is necessary

to appropriately reduce the weights of indicators that

exhibit high levels of correlation.

4) Computational comprehensive information

The objective weight of each index is

comprehensively measured by its variability and

conflict. Let

j

C

represent the comprehensive

information contained in the

jth evaluation index,

then

j

C

can be expressed as:

1

*(1 ) *

n

jj ij jj

i

CrR









(14)

The larger

j

C

is the greater the role of the

th

j

evaluation index is, and the more weight should be

assigned to it.

5) Calculated objective weight

Therefore, the calculation formula of the

jth

index is:

1

j

CRITIC

n

j

i

C

W

C





(15)

(3) A combination calculation method of

subjective and objective weight based on

AHP-CRITIC method

The subjective weighting method of AHP mainly

relies on the subjective judgment of evaluators and

lacks reliability and stability. The objective weight

method of CRITIC just judges the importance of

indicator from sample data, without taking into

account information other than data, and the sample

data itself has certain limitations. Each of these two

methods has its own advantages and disadvantages.

Therefore, in this paper, they are combined. First,

AHP method and CRITIC method are used to

calculate indicator weight respectively, and then

weight data are combined to calculate combined

weight, so as to carry out performance evaluation on

Research on Post Evaluation Method for Power Grid Technology Projects Based on AHP-CRITIC-TOPSIS

267

this basis. The specific combination weight

calculation formula is as follows:

1

*

jj

AHP CRITIC

j

n

AHP CRITIC

j

WW

W

WW





(17)

3.2 TOPSIS Evaluation Methods

TOPSIS method is a classic data-driven evaluation

method. It was put forward by C. L. Hwang and K.

Yoon in 1981, ranking the proximity between

evaluation objects and idealized targets to determine

their relative merits and demerits.

In this paper, a multi-objective comprehensive

evaluation method combining CRITIC method, AHP

combined weight method and TOPSIS method is

adopted to analyze power grid science and technology

projects.

The specific calculation process of TOPSIS

method is shown as follows:

(1) The original matrix is turned forward. The so-

called positive transformation means that all types of

indicators are converted into extremely large

indicators. The conversion process of different types

of indicators is also different. The specific conversion

process is as follows:

1) Very small indicators—> very large indicators

'

x

max x

(17)

Where,

'

x

is the transformed index value,

x

is the

extremely small index value, and

max

is the

maximum value of this extremely small index in all

evaluation objects.

2) Intermediate indicators—> very large indicators

max{| |}

ibest

Mxx (18)

'

||

1

ibest

i

xx

x

M





(19)

Where,

'

i

x

is the index value after transformation,

i

x

is the intermediate index, and

best

x

is the best value

in the index value.

3) Interval type indicator—> extremely large

indicator

max{ min{ }, max{ } }

ii

M

ax xb  (20)

'

1,

i

ii

i

ax

x

a

M

x

ax b

xb

x

b

M























(21)

Where,

'

i

x

is the index value after transformation,

i

x

is the interval type index value, and [,]ab is the

interval of the index.

(2) The forward matrix is normalized to eliminate

the influence of different index dimensions.

Assuming that there are

n

objects to be evaluated

and m evaluation indicators after normalization, the

forward matrix formed is shown in Equation (22)

below.

11 12 1

21 22 2

12

n

mm mn

x

xx

x

xx

X

x

xx

















(22)

Where,

ij

x

represents the index value of the

ith

index of the

jth evaluation object.

The normalized matrix is denoted as

Z

, then the

calculation formula of element

ij

z

in matrix

Z

is

shown in Equation (23) below.

2

1

ij

m

ij

j

x

z

x





(23)

(3) Calculate the final score.

Assuming that there are

n

objects to be evaluated

and

m evaluation indicators, the final standardized

matrix is obtained, as shown in Equation (24) below.

11 12 1

21 22 2

12

n

mm mn

zz z

Z

zz z

















(24)

Define maximum

Z



:

12

11 21 1 12 22 2 1 2

(, ,, )

(max{ , , }, max{ , , }, , max{ , , })

n

mmnnmn

ZZZ Z

zz z zz z zz z

 







(25)

Define minimum

Z



:

12

11 21 1 12 22 2 1 2

(, ,, )

(min{ , , }, min{ , , }, , min{ , , })

n

mmnnmn

ZZZ Z

zz z zz z zz z

 







(26)

Define the distance

j

D



between the

1, 2, ,jth, j n



… evaluation object and the maximum

value.

2

1

()

m

jiiij

i

D

wz z









(27)

ANIT 2023 - The International Seminar on Artiﬁcial Intelligence, Networking and Information Technology

268

Define the distance

j

D



between the

1, 2, ,jth, j n … evaluation object and the minimum

value

2

1

()

m

jiiij

i

Dwzz









(28)

In the final calculation, the score

j

S

of the

1, 2, ,jth, j n … rating object without normalization

can be written:

j

D

S

DD









(29)

Obviously, and the higher the score is

01

i

S



 ,

the larger

j

D



is, that is, the closer it is to the

maximum value.

（4) The scoring criteria are normalized



1

j

n

j

S





(30)

Where,



j

S is the score of the jth evaluation

object after normalization, Obviously,



1

n

j

S





.

4 CASE ANALYSIS

(1) Determine subjective weight based on AHP

Taking the calculation of the first-level indicator

layer as an example, the judgment matrix of the first-

level indicator layer is determined through expert

review and scoring, as shown in TABLE 4 below.

Table 4. First-level indicator judgment matrix.

M

1

A

2

A

3

A

4

A

1

A

1 1/3 1/5 1

2

A

3 1 1/3 3

3

A

5 3 1 5

4

A

1 1/3 1/5 1

Through the calculation and analysis of the above

matrix, the weight of the final index layer is

determined as: 0.0976, 0.2516, 0.5549, 0.0967,

0.0056 0.1CR , the matrix passes the consistency

test.

Similarly, the index weight of the second-level

indicator layer relative to the first-level indicator layer

is further determined, and the final weight of each

indicator is obtained as shown in TABLE 5 below.

Table 5. Indicator weights based on AHP.

Indicator Weight Indicator Weight

1

B

0.0242

5

B

0.1387

2

B

0.0725

6

B

0.4162

3

B

0.1258

7

B

0.0725

4

B

0.1258

8

B

0.0242

(2) Determine objective weight based on

CRITIC

Based on the actual data of each science and

technology project and the basic theory of CRITIC,

the results of the weight of each index are calculated,

and the objective weight results are shown in TABLE

6 below.

Table 6. Indicator weight based on CRITIC.

Indicator Weight Indicator Weight

1

B

0.1081

5

B

0.1301

2

B

0.1992

6

B

0.0819

3

B

0.1138

7

B

0.1295

4

B

0.1382

8

B

0.0992

(3) Combinatorial weighting

Based on the above equation (16), the subjective

and objective combination weights are calculated, and

the final weights of each index are obtained as shown

in TABLE 7 below.

Table 7. Indicator combination weighting.

Indicator Weight Indicator Weight

1

B

0.0232

5

B

0.1602

2

B

0.1282

6

B

0.3024

3

B

0.1271

7

B

0.0834

4

B

0.1543

8

B

0.0213

(4) Comprehensive evaluation based on TOPSIS

Based on the theory of TOPSIS method, the

distance between the object to be evaluated and the

positive ideal point and the negative ideal point is

calculated, and the final evaluation results are given,

as shown in TABLE 8 below.

Research on Post Evaluation Method for Power Grid Technology Projects Based on AHP-CRITIC-TOPSIS

269

Table 8. Scores of the objects to be evaluated.

Project

j

D



j

D



j

S



j

S

Big data

scheduling technology

0.0501 0.0335 0.4009 0.1288

Virtual power

plant technology

0.0325 0.0496 0.6043 0.1941

DC networking

technology

0.0349 0.0454 0.5651 0.1815

DC microgrid

technology

0.0357 0.0410 0.5347 0.1718

AC-DC hybrid distribution

network technology

0.0342 0.0364 0.5156 0.1656

Digital compound

networking technology

0.0473 0.0459 0.4922 0.1581

As can be seen from the above table, among the six

technologies of big data scheduling technology,

virtual power plant technology, DC networking

technology, DC microgrid technology, AC-DC hybrid

distribution network technology, and digital

compound networking technology, the order of

evaluation is virtual power plant technology, DC

networking technology, DC microgrid technology,

AC-DC hybrid distribution network technology,

digital compound networking technology and big data

scheduling technology.

5 CONCLUSION

This paper investigates the indicator system and

evaluation model for post-evaluation of technology

projects within power grid companies. Firstly, based

on the relevant concepts of post-evaluation for

technology projects, the post- evaluation objectives

for power grid technology projects were clarified, and

an indicator system for post-evaluation of technology

projects was constructed. Then, by considering the

characteristics of these indicators, a technology

project post-evaluation model based on AHP-

CRITIC-TOPSIS was developed, and typical power

grid technical projects were evaluated by collecting

data. Feasibility of this method was validated through

a case analysis.

ACKNOWLEDGMENT

This study is supported by the project of State Grid

Economic and Technological Research Institute Co.,

LTD. “Technology maturity evaluation and post-

project evaluation of key technologies and application

research” (SGSH0000SCJS2200105). The authors are

grateful to the participants who help to improve the

paper with many pertinent comments and suggestions.

REFERENCES

Grid modernization multi-year program plan(R). U. S.

Department of Energy, 2015.

Amin D, L C T. A complex network approach for analyzing

early evolution of smart grid innovations in Europe (J).

Applied Energy, 2021, 298: 117143.

Ghizoni C R T, Roesler L A R, Paulo J P. An Intelligent

System for Operators Performance Multicriteria

Evaluation of Distribution Networks (J). Applied

Artificial Intelligence, 2022, 36(1): 1-24.

Areeba N, Kifayat U, Maria A, et al. Assessment of Smart

Grid Systems for Electricity Using Power Maclaurin

Symmetric Mean Operators Based on T-Spherical

Fuzzy Information (J). Energies, 2022, 15(21): 1-25.

Xiufan M, Ying W, Zihao L, et al. Comprehensive

evaluation of 5G+ smart distribution network based on

combined weighting method-cloud model(J). Global

Energy Interconnection, 2022, 5(6): 675-691.

Long C W, Han F, Zhang M X, et al. The Relationship

among the Security, Efficiency and Adequacy of

Electrical Power Grid Based on the Novel Evaluation

Method (J). Advanced Materials Research, 2014, 3247:

960-961.

Hu Z, Hu L, Cheng H, et al. Application of Pollution Index

Method Based on Dynamic Combination Weight to

Water Quality Evaluation(J). IOP Conference Series:

Earth and Environmental Science, 2018, 153(6): 1-3.

ZiWei L, QiongYao X, Lu D, et al. Research on

comprehensive evaluation method of distribution

network based on AHP-entropy weighting method (J).

Frontiers in Energy Research, 2022, 10: 1-9.

ANIT 2023 - The International Seminar on Artiﬁcial Intelligence, Networking and Information Technology

270