You are on page 1of 4

Topsis-AHP-Simulation Method and Its Application in Operational

Capability Evaluation
Yin Chun1, Huang Yan-yan1,Wang Zhi-quan1
1. Automation School, Nanjing University of Science and Technology, Nanjing ,210094
E-mail:Hyy30004000@163.com

Abstract: It is necessary to evaluate the operational capability of the weapon equipment based on combat-simulation, however, the
simulation data is so isomerous that it is difficult to make evaluation model. Pointing to this problem, a comprehensive method based on the
TOPSIS, AHP and SIMULATION method (T-A-S) is built to evaluate the operational capability of weapon equipment. An example about the
operational capability on armored equipment based on the T-A-S method is given, and the evaluation results can help to analyze the weapon
combating performance. The results show the comprehensive method very feasible.

Key Words: Evaluation method, Simulation, Operational capability

decision-making method, and we can measure whether an


1 INTRODUCTION alternative good or not on the Euler distance which is from
In the demonstration of weapon equipment, it is necessary an alternative’s attributes to the ideal alternative attributes
to use the simulation method to evaluate the operational point. Euler distance is widely used in the T-A-S method to
capability (OC) of the weapon equipment [1]. However, the measure the operational capability of the weapon
evaluation data from the simulation method is isomerious, equipment.
and the data difficult to integrate into a comprehensive As we know, the common weapon equipment has such
evaluation value. In fact, there are two main kinds of the essential capabilities such as the firepower capability,
data to evaluate the operational capabilities: operational data moving capability, protection capability and the
and the un-operational data. The operational data is from the communication capability and so on. In this paper, we build
combat-simulation, and the un-operational data from the four indexes systems for the essential operational
experts’ experience. The un-operational capability data is capabilities of weapon equipments. And these four indexes
easy to get but difficult to integrate into an evaluation value. systems can be used to evaluate the capabilities
And therefore, we need use two kinds of evaluation data to independently. According to the requirements of the
discuss the weapon equipment operational capability. The equipment evaluation, the evaluation alternatives we
reference [2] uses the TOPSIS (Technique for Order designed have many factors, and for each indexes system of
Preference by Similarity to Ideal Solution) method to operational capability, adopt the AHP method to get the
evaluate the equipment alternatives. In fact, the equipment indexes weights, and use simulation method to get the
capability evaluation can be seemed as the evaluation of evaluation data, use the TOPSIS method to get the
equipment alternatives. In this paper, in order to research the comprehensive evaluation of every alternative. In essence,
operational capability of weapon equipment, the methods we use the Euler distance to measure the operational
such as the TOPSIS [3],AHP[4] and the simulation capability of the weapon equipment.
method[5] are used to make the comprehensive evaluation
model. 2.2 AHP Method and Steps

2 TOPSIS- AHP -SIMULATION METHODS When the indexes system of the operational capabilities is
created (see in the Fig.1), we can use AHP method to
2.1 T-A-S Methods compute the indexes weights value.
Ca pa bility A
T-A-S is the combination of three methods: TOPSIS,
AHP and the Simulation. It is a comprehensive evaluation
Sub Sub Sub Sub
framework, which includes simulation data collection, c apa bil ity ca pab ili ty ca pa bil it y c apa bi lit y
of A : A 11 A 12 A 1i A 1k
expert judgment, and system evaluation method. Firstly, it
needs an indexes system to describe the operational
capability, and then get the indexes weights through AHP Sub of S ub of Sub of Su b of Su b of Su b of
method; secondly, get the indexes values by means of A 11 : A 11 : A 12 : A 1 i:
A2j
A 1 k:
A2h
A 1k :
A 2n
A 21 A 22 A 23
system simulation; finally, use the TOPSIS method to
comprehensively evaluate the weapon equipment
Fig.1: The capability indexes system of evaluation
alternatives. In this evaluation framework, the TOPSIS
method is the core of the model. It is a multi-attribute The indexes weights are important to evaluate their
operational capabilities. According to the indexes systems,

978-1-4244-2723-9/09/$25.00 
c 2009 IEEE 2954
the steps to compute the indexes weights by the AHP as developers. The simulation data list describes the detailed
follows: information about the combat capabilities of weapon
1)create the judgment matrix based on the characteristics of equipments.
the equipment operational capability. During the simulation process, we use the XML files to
In the hierarchy of indexes systems of the equipment transfer and save the simulation result data. The simulation
operational capabilities, the matrix is described by the data file we collected are very large and complex, but we can
importance degree of a layer indexes contrasting with their easily deal with the data by means of the XML language.
upper layer index. For the index Ak ˉ 1 ˈ m , it has a But the XML file form (DTD) of the simulation data list
relationship with its sub-indexes like Akˈ,j. By the means of must agree with the form of the evaluation requirement file.
When a simulation alternative is finished, the raw
the Saaty’s 9-scale method, the judgment matrix is easy to
make. simulation data are saved by the XML file in the given form
of simulation data collection list. In the same way, running
Tab. 1: Judgment Matrix
many simulation alternative, and get the simulation data
Ak Ak1 Ak2 … Ak n
files respectively, which comes from a different simulation
Ak1 b11 b12 … b1n
Ak2 b21 b22 … b2n
alternative. The files have much information to evaluate the
… … … bij …
alternatives of weapon equipments.
Akn bn1 bn2 … bnn Usually, it is uneasy to deal with the higher level indexes
data from the traditional simulation, because the data we
In the Table 1. The scale sets {bi j } ∈ {1, 2,3,...,9} .
gotten are such a raw data that we need make a data
2) multiply the elements of each row for the judgment pre-processing tool. The pre-processing work mainly
matrix, and the result is Mi. includes the statistical calculation and the raw data merger.
n After the raw data are processed into the standard data type,
M i = ∏ bij ,i˙1,2,…,n. (1) we can easily evaluate them. And therefore, a
j =1
pre-processing module is developed. With this
3) calculate the n-th root of Mi pre-processing tool, we can transfer smoothly the raw
−−
simulation data to the standard data for evaluation, and the
W = n Mi
standard data are transferred and saved in form of XML
−− −− −− −−
W = [W1 W2 ... Wn]
files.
4) Standardize the eigenvector of . It is
the needed eigenvector, and it is the weight vector. 2.4 TOPSIS Method
− n −
TOPSIS is the name of Technique for Order Preference
W i
=Wi ¦Wj =1
i (2)
by Similarity to Ideal Solution for short. It measures the
5) calculate the maximal the Eigen value λmax from the judgment multi-attributes alternatives with the Euler’s distance.
matrix as follows: As we know, the operational capability of an equipment
n
system is evaluated by many indexes or attributes. Most of
( AW )i
λmax = ¦ (3) the attributes values in this paper are collected from the
i =1 nWi simulation. To evaluate the equipment capabilities, the
6) test the coincidence indicator and make the rank evaluation matrix is abstracted from the detailed simulation
reasonable data. It is an alternative-attribute matrix, and the values of
coincidence indicator (CI) the attributes come from the simulation results data.
CI˙( λmax ˉn)/(n-1) (4) Tab. 3: Alternative-attribute List
alternatives Attri 1 Attri 2 … Attri n
Average coincidence indicator of random (CR)
alternative 1 C11 C21 … C1n
CR=CI/RI<0.1 (5)
… … … FLM …
RI is the same rank CRˈshown in Tab.2. If the weight
alternative n Cn1 C22 … Cnn
values are not agreed with the judgment matrix, it needs to
pass the coincidence test. In the table 3, cij means that the jth attribute value of the ith
Tab.2: Average Coincidence Indicators of Random alternative, which can get from the simulation results. The
Rank 1~2 3 4 5 6 7 8 9 steps of TOPSIS evaluation model are as following:
R.I. 0 0.52 0.89 0.12 1.26 1.36 1.41 1.46 1) transfer the attributes value from Tab. 3 into the initial
evaluation matrix C.
2.3 Simulation Data Collection C˙ §¨ cc 11 c 12 ... c1n ·
¸
¨ 21 c 22 ... c2n ¸
In the equipment demonstration, the operational ¨ ... ... ... ... ¸
¨ ¸
capability evaluation is the final aim of combat-simulation. ¨ c c ... c mn ¸¹
© m1 m 2
So it is necessary to build the indexes system at beginning,
so as to detail what simulation data to be collected. Considering of the attributes values in horizontal list of
According to the evaluation requirements, build a matrix C standing for different indexes, they can not be
simulation data list to guide the modeler and simulator to get contrasted with each other, but those which come from
the very engagement result data of the weapon equipment. vertical list of matrix C can be compared with each other
And so, it is important to create such a simulation data list because the vertical list values are in the same index form.
for the both the clients and the modeling &simulation With that, standardize evaluation matrix C and get its

2009 Chinese Control and Decision Conference (CCDC 2009) 2955


standard matrix R. the attribute value rij is denoted in the For short, in the paper, only gives the D1. And we suppose
formula 6. The standard matrix R is essential to the get maximal Eigen value of D1 is 9.0441. By means of
following evaluation steps. formula (4) (5), get the CI=0.0055125, and check up the
n table 3 to get the RI=1.46.At last, get the C.I.=0.00376<0.1,
(6)
ri j = w j ∗ c i j / ci =1
2
ij and this condition means the D1 having the coincidence. The
similar way is done to the D2, D3.
2) seek the most positive or negative alternative from R. And therefore, we can get the weight-vectors W of all the
To seek the most positive or the worst negative firepower capability matrix, W˙
alternative, we need distinguish the benefit-type attribute
(0.112,0.112,0.054,0.104,0.104,0.058,0.201,0.172,0.083˅
from the cost-type one, and take different way to deal with
them. The best positive alternative x* is denoted in formula Secondly, collect the raw evaluation data through the
7. simulation system.
{(max rij | j ∈ J ), (min rij | j ∈ J ') | i ∈ M } = [ r1* , r2* ,..., rn* ] (7) We select the i-th equipment alternative as a simulation
i i
experience. And the simulation data of the firepower
In the same way, the worst negative alternative xˉ is capability of the i-th equipment alternative is described in
denoted in formula 8˖ form of XML language as follows:
{(min rij | j ∈ J ), (max rij | j ∈ J ') | i ∈ M } = [ r1− , r2− ,..., rn− ] (8) …<firepower data> <times of hitting on object: shot at
i i MBtank="5" shot at anti-tank missile vehicle ="8" shot at
In above formulas, the M represents the alternatives set, the artillery object="9"/>
and J and J’ stand for the benefit type and cost type of <Average distance of shooting>
attributes sets. <average distance of hitting on MBtank: original location
3) The Euler Distances of each alternative to x* , xˉ. x="3" y="3" falling point location x="4" y ="4.5"/> …..
The Euler Distances of the i-alternative away to x*: < Average distance of shooting at the artillery: original
location x ="4" y="4" falling point location x ="5.4"
(i ∈ M)
n

S
*
i
=  ( ri j
− rj )
* 2
(9) y="5.5"/> </ firepower data >…
j = 1
Thirdly, deal with the original simulation data, and get the
the Euler Distances of the i- alternative away to xˉ˖ basic indexes data to build the evaluation matrix.
For instance, when we think about the equipment
(i ∈ M)
n

S i

=  ( ri j
− rj )
− 2
(10)
alternative 1, the important thing is to collect the simulation
j = 1

data, deal with them and fill the very simulation data into the
4˅calculate the approaching degree of the ith alternative evaluation matrix based on its index attributes. And the
approach to the most positive alternative other alternatives are similar to the alternative 1. And
− −
A i
= S i
/( S i
+ S *
i
) (11) therefore, we get evaluation matrix C about operational
The more of the approaching degree of the alternative is, capability as follows:
the best of the alternative does. We can use the Ai to describe 5 8 9 1.80 1.92 2.05 2 3 8

the capability of the equipment and make a rank. C˙ 6 7 9 1.90 1.85 2.20 3 3 7 
9 8 8 1.98 1.89 2.02 5 3 5

3. Application example 7 9 7 2.00 1.90 2.10 3 4 4

Fourthly, uniform the evaluation matrix to D.


In order to evaluate the operational capability of weapon
D=
equipments, we decompose the operational capabilities into
four basic parts: firepower, maneuverability, protection, 0.362 0.498 0.543 0.468 0.508 0.490 0.292 0.457 0.535

communication. As an example, the firepower capability of 0.434 0.434 0.543 0.494 0.489 0.525 0.438 0.457 0.624
the armored equipment is evaluated in the T-A-S method. 0.651 0.498 0.482 0.515 0.50 0.482 0.729 0.457 0.445 

Suppose that there are four different equipments as follows:  0.507 0.560 0.422 0.520 0.503 0.502 0.438 0.610 0.356 
first alternative is to load new armored, second one is Fifthly, evaluate the operational capability of each
improved the equipment marching speed, the third alternative in form of the Euler distance. The evaluation
alternative improves collimator, the fourth alternative value of firepower capabilities of equipment alternatives are
employ advanced transmitter-receiver. The detail evaluation following.
steps as follows:
0.041 0.056 0.029 0.049 0.053 0.028 0.059 0.079 0.044
Firstly, build up the indexes system of the firepower 
capability, and use AHP method to calculate the indexes 0.049 0.049 0.029 0.051 0.051 0.030 0.088 0.079 0.052
0.073 0.056 0.026 0.053 0.052 0.028 0.147 0.079 0.037 
weights. 
In view of the firepower capability A and its sub indexes  0.057 0.063 0.023 0.054 0.052 0.029 0.088 0.105 0.030 
such as A11, A12, A13 and so on, we can build the judgment Finally, the evaluation values of the operational capabilities
matrix D. for 4 equipment alternatives are:0.4937, 0.4989, 0.5063,
 1 1 2  0.5000 .
D =
 1 1 2 
 From the above firepower capabilities, we can easily
 1 / 2 1 / 2 1 

know that the 3rd alternative is the best, and the 4-th
For the sub indexes A11,A12,A13 and their sub indexes, we alternative is the second. At the same time, we can contrast
can make the three judgment matrix as follows:D1,D2,D3. the factors of the alternatives with their evaluation results,

2956 2009 Chinese Control and Decision Conference (CCDC 2009)


and we can draw a conclusion that the 3rd alternative is good Weapon and Equipment[M], Beijing:National Defense
because it improves the precision of collimator tools. Industry Public , 2000
According to the conclusion and the advice, the design and [2] Ma Y.L, Wang JY, Xu XZ, Study of Evaluation Method for
revision of weapon equipments have a clear direction so as Simulation Result Based on Positive and Negative Ideal
Point[J], Computer Engineering, 2002.(2):21-22.
to improve the weapon equipment performance. [3] Huang Y.Y; Yang F; Liu C; Simulation Based Evaluation
Similarly, with the T-A-S method, we can do the same Framework of Weapon Equipment Supporting Traceability
thing for the other capability alternatives such as and Its Application[J], Journal of System
maneuverability, protection, communication and so on. simulation,2006.(5): 1360-1365
[4] Cheng Qin-wen; Shen Yun-chun; Tang Chun-jie, Research on
the Synthetic Evaluation Model of Warfare Capability of
4. Conclusion Submarine Launched Torpedoes, Fire Control & Command
In this paper, the comprehensive evaluation framework Control, 2004,29(2):109-111
[5] Huang Y.Y; Yang F; Wang WP; Research of Evaluation
including the AHP, TOPSIS and simulation method is
Methodology of Armored Equipment Operational
introduced to evaluate the operational capabilities of Effectiveness Based on Utility Fusion, Journal System
weapon equipment. With the help of the TAS method, we simulation, 2005.(10): 2311-2313
can easily evaluate the weapon equipments alternatives, and
can find out the key factor which affecting the weapon
performance. An example about the armored equipments Writers introduce:
evaluation is carried out by above methods, and the results YIN Chun(1962,01-),male, born in Pingxiang city, Jiangxi province,
postgraduate, major in: systems engineering
show reasonable. HUANG Yan-yan˄1973ˈ10ˉ˅ˈborn in Lipu county, Guangxi province,
PH.D degree, major in: systems engineering, modeling and simulation,
command control and automation

5. REFERENCES First writer Link information:


Address: Automation School of Nanjing University of Science and Technology,
[1] Li Ming, Liu Peng, Theory and Practice of an Integrated and Nanjing ,Jiangsu province, china ,210094
Systematic Demonstration method for Development of

2009 Chinese Control and Decision Conference (CCDC 2009) 2957

You might also like