Ain Shams Engineering Journal 14 (2023) 102055

Contents lists available at ScienceDirect

Ain Shams Engineering Journal

journal homepage: www.sciencedirect.com

Artificial intelligent applications for estimating flow network reliability

Moatamed Refaat Hassan a, Salem Alkhalaf b, Ashraf Mohamed Hemeida c,⇑, Mahrous Ahmed d,
Eman Mahmoud e
a
Computer Sciences Branch, Department of Mathematics, Faculty of Sciences, Aswan University, 81528 Aswan, Egypt
b
Department of Computer, College of Science and Arts in Ar-Rass, Qassim University, Ar Rass 52571, Saudi Arabia
c
Department of Electrical Engineering, Faculty of Energy Engineering, Aswan University, 81528 Aswan, Egypt
d
Electrical Engineering Department, College of Engineering, Taif University, Taif 21944, Saudi Arabia
e
Department of Mathematics, Faculty of Sciences, Aswan University, 81528 Aswan, Egypt

a r t i c l e i n f o a b s t r a c t

Article history: Artificial intelligence (AI), often known as machine learning, is a powerful tool for solving engineering
Received 8 October 2022 problems. The evaluation of the network reliability of a flow network is a NP-hard problem, with compu-
Revised 14 November 2022 tational effort growing exponentially with the number of nodes and arcs in the network. Also, the com-
Accepted 27 November 2022
ponents assignment issue is NP-hard, and the computational effort increases with the number of
Available online 09 December 2022
available components. Many candidate solutions are typically examined during optimal components or
optimal capacity assignment, each requiring reliability calculation. Consequently, this paper proposes
Keywords:
an artificial neural network (ANN) predictive model to evaluate the flow network reliability. The neural
Artificial intelligent
Flow network
network is one of the artificial intelligence tools constructed, trained, and validated using the maximum
Components assignment problem capacity of each component input and the network reliability as the target. The proposed ANN model pro-
Capacity assignment problem vides empirical proof that neural networks can accurately estimate reliability by modeling the connection
Reliability evaluation between the maximum capacities of network components and the reliability value.
ANN Ó 2022 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams Uni-
versity. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

1. Introduction ability at each level and the reliability of such a flow network can
be evaluated as a function of K-lattices generated from each subset
Practical systems such as energy transmissions through trans- of the family of all MCs (minimal cutsets). However, the size of
mission, distribution, and communication systems can be consid- such a family 2 m-1 (m = number of MCs) grows exponentially
ered flow networks whose parameters are independent, vague, and with m. A flow network of this type can be considered a multistate
multivalued random capacities. A Flow network is known as many system with multistate components. Its reliability can be mea-
states approach with multistate components, hence its reliability sured in terms of the upper boundary points of each level d.
for the system demand, d. The probability that the maximal flow is (named d-MCs here). The generation of all d-MCs from each MC
not less than d can be calculated from minimal path vectors to level for each system capacity level d was discussed and developed [3].
d (named d-MPs here). A simple algorithm to create all d-MPs of such The components assignment problem is the difficulty of finding
a system for each system capacity level d in terms of minimal path the best components that enhance reliability while minimizing
sets was investigated and discussed [1]. There are three main issues total cost or lead time [5–14]. A genetic algorithm (GA) is applied
related to a flow network; reliability evaluation, components assign- successfully to solve stochastic problems traditionally associated
ment problem, and capacity assignment problem. with Operations Research (OR). GA produced high-quality results
The reliability of a flow network has been evaluated in terms of for solving the generalized assignment problem than traditional
minimal paths [2] or minimal cuts [3–4]. The maximum flow prob- OR heuristic approaches [5]. The network reliability under compo-
nents assignment can be estimated as a function of minimal paths
⇑ Corresponding author. and state-space decomposition. GA is successfully implemented to
E-mail address: [email protected] (A. Mohamed Hemeida).
solve the network reliability assignment problem quickly and
Peer review under responsibility of Ain Shams University. accurately. [8]. Multistate components assignment problem with
optimal network reliability subject to assignment budget
restriction, where each component has an assignment cost, is
investigated and discussed. The problem is assessed using minimal
Production and hosting by Elsevier paths and state-space decomposition [9]. Maximizing network

https://doi.org/10.1016/j.asej.2022.102055
2090-4479/Ó 2022 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams University.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

reliability for stochastic transmission networks is discussed and The maximum capacity of component (arc or node) is an essen-
evaluated based on GA [10]. tially to deduce the component’s information (no. of states, capaci-
An optimization algorithm integrating a genetic algorithm, min- ties, and probabilities). So, it plays an important role in evaluating
imal paths, and the Recursive Sum of Disjoint Products (RSDP) is network reliability. Consequently, this paper implements an artifi-
suggested to solve such a problem. A practical case of LCD TV from cial neural network (ANN) model to predict the reliability of a flow
China or Germany is a real example to investigate the solution pro- network based on the maximum capacities of its components. The
cedures. An exact optimal double-resource assignment for the proposed ANN is trained based on maximum capacities of the net-
robust design problem in multistate computer networks is investi- work’s components (arcs or nodes). The reliability as a function of
gated. The networks have links and vertices where both kinds of components’ capacities is performed by using the proposed ANN
resources may have several states due to failure or maintenance. model. The feature of ANN is more efficient than the NEA (Nonlinear
An efficient approach is suggested for such an assignment and extension of Arps decline). Also, it provides better and accurate
illustrates the technique’s efficiency using numerical examples results.
[11]. System reliability maximizes a computer network by calcu- The paper is orderly as follows: Section 2 presents reliability
lating the accurate two-class allocation subject to budget [12]. evaluation. Section 3 describe how to generate arcs’ information.
Solving a component assignment problem for a stochastic flow net- In section 4, we give the proposed ANN components. Section 5 clar-
work under lead-time constraint [13] and a two-stage approach for ifies training and testing processes. The discussion is given in sec-
a multi-objective component assignment problem for a stochastic- tion 6. The last section presents the conclusion.
flow network [14] are investigated.
An optimal capacity assignment for the robust design problem 2. Reliability evaluation of a flow network
in capacitated flow networks is investigated and discussed. Opti-
mal resource allocation of computer network nodes provides an The reliability of a stochastic flow network (SFN) for a given
improvement in system reliability. This approach is focused on demand d (Rd ) is evaluated by the following equation:
searching the minimum capacity assignment of each edge such
that the network still survived even under the edge’s failure [15].
Rd ¼ PrfXjVðXÞ  dg ð1Þ
Optimal resource allocation of computer network nodes provides Where VðXÞ is the maximum flow under a lower boundary vec-
an improvement in system reliability. Capacity assignment tor X. The capacity vector X ¼ ðx1 ; x2 ; x3 ;    ; xm Þ can be deduced
approaches are investigated by analyzing the node structure with from the flow vector F ¼ ðf 1 ; f 2 ; f 3 ;    ; f np Þ by the following
respect to the system reliability of a stochastic-flow network relation:
[16]. The capacity assignment problem, on the other hand, requires
finding the optimal capacity for each component (arc or node) in a X
np
 
xi ¼ f j jai 2 mpj for each i ¼ 1; 2;    ; m ð2Þ
way that simultaneously minimizes total capacities while maxi- j¼1
mizing reliability [17–19]. Effective solving of robust design prob-
lems associated with capacity allocation in stochastic-flow Where m and np is the number of arcs and minimal paths
networks was investigated and discussed [20]. During the process respectively.
of selecting the optimal components or capabilities, a large number Finally, the arc’s information (number of states, capacity, and
of possible solutions are explored, each of which requires reliabil- probability of each state) should be known in advance to evaluate
ity calculation. Furthermore, the assignment and network reliabil- Rd .
ity evaluation problems are NP-hard [21,22]. Let X 1 ; X 2 ; X 3 ;    ; X q be the obtained lower vectors, then Rd is
ANN is a computational approach that maps the non-linear expressed as follows:
relationship between input and output variables based on the n[q n oo
i
training process. Predicting using ANNs is applied to many compli- Rd ¼ Pr i¼1
XjX  X ð3Þ
cated problems in engineering and science [23–31]. In [25], ANN
predicts the PCI (Pavement condition index). An Adaptive PrfXg ¼ Prfx1 g  Prfx2 g  Prfx3 g      Prfxm g
Network-based Fuzzy Inference System (ANFIS) and ANN-based
Eq. (3) can be efficiently calculated by using one of the best-
strategy were proposed by [24] to forecast the tensile strength of
known methods, state-space decomposition [42], inclusion–exclu-
polyurethane wood ash composite. Oil and natural gas forecasting
sion [43] or quick inclusion–exclusion [44] or RSDP [45].
is another area where ANN has proven useful [30,32].
So, the reliability evaluation of a flow network is an NP-Hard
An effective method for predicting obsolescence based on a
problem.
modified radial basis neural network was analyzed and investi-
gated to locate the clustering centroid [33]. Multiple linear regres-
sion and ANN for water prediction in Egypt were developed [34]. A 3. Generating Arc’s information
novel switched predictive model control of wind turbines using
ANN prediction with load mitigation was discussed and investi- Let the maximum capacity of an arc ai is M i . Assume that the
gated [35]. Convolution neural networks (CNN) have proven high current state and the capacity of ai is sj and cj respectively. The cor-
accuracy implementation of urban building footprints [36]. The responding probability pj of the state sj can be deduced by using
ANN approach was implemented successfully and accurately for the following relation.
predicting the pavement condition index [37]. Feed-forward neural !
  M i sj i
network classifiers based on nature-inspired algorithms were dis- pj ¼ Pr sj ¼ r ð1  r i ÞM sj ð4Þ
cussed and investigated for one decade [38]. Higher order neural sj i
network was implemented for chemical reaction optimization for
Where, ri is reliability of an arc ai .
non-linear classification investigation [39]. Artificial chemical
For example, let the maximum capacity of a1 is 2, then the
reaction optimization of neural networks for efficient prediction
states (capacities) are 0, 1, 2. The probabilities are:
of stock market indices was discussed and investigated [40]. An  
online handwritten signature verification system based on Discrete 2
p0 ¼ ð0:90Þ0 ð1  0:90Þ20 ¼ 0:01
Wavelet Transform (DWT) features extraction, and neural network 0
classification was discussed and developed [41].
2
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

 
2
p1 ¼ ð0:90Þ1 ð1  0:90Þ21 ¼ 0:18
1
 
2
p1 ¼ ð0:90Þ2 ð1  0:90Þ22 ¼ 0:81
2
Noting that the total probabilities, p0 þ p0 þ p0 ¼ 0:01þ
0:18 þ 0:81 ¼ 1.
Fig. 2. Six-arc network.

4. The proposed ANN model

The FNN is suitable for the presented study. The input item con-
tains the set of maximum capacities and the corresponding relia-
bility for training.
The following subsections explain the structure of the proposed
FRNN model.

4.1. Data scaling

The data should be scaled to be within the scale of the activa-

tion function used by the NN.

4.2. Scale-dependent error

The RMSE (Root Mean Square Error) is preferred and popular

because of their theoretical relevance in statistical modeling, [35] Fig. 3. Fourteen-arc network.
and [42].
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u n
u1 X 2
RMSE ¼ t ðRsobs i  Rspred i Þ ð5Þ
n i¼1

4.3. Percentage error

The most commonly used measure is the root mean square per-
centage error (RMSPE) [35] and [42], which can be given as follows:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 0 1
u X obs 2
u1 n Rspred  Rs
RMSPE ¼ t @ i i A
 100 ð6Þ
n i¼1 Rsobsi

4.4. Mean Absolute Error

Mean Absolute Error (MAE) is given by [35] and [42]:

1X n
MAE ¼ jRsobs
i  Rspred
i j ð7Þ
n i¼1

4.5. Reference model

FRNN model is preferred because of less errors, [46]. Fig. 1

shows a simple ANN with set of inputs and a single output.

Fig. 1. ANN network. Fig. 4. Thirty-three-arc network.

3
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Fig. 5. Testing results AD for six-arc network. Fig. 7. Testing results AD for Thirty-three-arc network.

5. Training and testing the ANN Table 1

Values for MSE, MAE, and RMSPE.

The network with four nodes and six arcs as shown in Fig. 2. The Error Six-Arc Fourteen-Arc Thirty tree-Arc
data used in training are shown in Table A.1, the total number \Network network network network
equals 1000; 800 for training and 200 for testing. The other net- MSE 0.00035 0.00031 0.00654
works with 14 and 33 arcs are shown in Figs. 3 and 4 respectively, MAE 0.01034 0.01233 0.06270
[18]. The data used in training and testing for other networks are RMSPE 3.38452 2.93338 15.78133

available on request. Figs. 5, 6, and 7 show the predicting results

in comparison with the actual results for testing data based on model can comprehensively predict reliability value within small
absolute difference (AD). While Table 1 shows the MSE, MAE, errors of the scaled dependent errors according to Table 1. In addi-
and RMSPE for each network. tion, the proposed ANN model can be used to predict the reliability
of a flow network for large size systems.

6. Discussion

The Absolute Difference (AD) was used here to show how pre- 7. Conclusions
dicted and exact values of reliability values close to each other.
Figs. 5, 6, and 7 summarize the AD values of testing data for each This paper presented ANN approach to estimating flow network
studied network. The lowest value of AD for Six-Arc network, reliability based the maximum capacities of components. The ANN
Fourteen-Arc network, and Thirty-tree-Arc network is 0.00026, trained and tested using different network examples. The recom-
0.00015, and 0.0 respectively. mended approach is to use the ANN during the optimal compo-
Using three scaled dependent errors (MSE, MAE, and RMSPE) to nents assignment problem to save computational efforts using
measure the accuracy of the proposed model in estimating net- existing methods or algorithms. The testing results assert the effi-
work radiality assert that the efficiency of it. The proposed ANN ciency of the proposed ANN in estimating reliability taking into
account the AD and scaled errors’ values.
The ANN approach implemented using Python and it can be used
within optimization algorithms. Finally, this approach can be
extended to cover all existing networks and widely used as an indi-
cator to the reliability of a flow network to save computational
efforts.

Declaration of Competing Interest

The authors declare that they have no known competing finan-

cial interests or personal relationships that could have appeared
to influence the work reported in this paper.

Acknowledgments

The authors would like to acknowledge the financial support

received from Taif University Researchers Supporting Project Num-
Fig. 6. Testing results AD for Fourteen-arc network. ber (TURSP-2020/146), Taif University, Taif, Saudi Arabia.
4
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Appendix

(See Table A1).

Table A1
The data used in training and testing for four-node example.

Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability
1 2 1 3 2 3 3 0.629032 334 2 2 4 4 4 3 0.920716 667 3 4 2 1 1 4 0.655346
2 3 3 2 2 3 3 0.971294 335 4 3 4 2 2 3 0.974601 668 3 3 2 1 1 4 0.653213
3 3 3 2 4 4 3 0.987403 336 3 3 2 1 3 2 0.924274 669 3 4 2 4 2 2 0.903232
4 2 3 4 3 3 3 0.889995 337 3 3 4 2 3 3 0.971365 670 3 3 2 2 1 2 0.621323
5 3 3 4 1 2 4 0.908911 338 3 3 4 3 3 4 0.978494 671 3 3 2 4 2 4 0.908474
6 2 3 2 2 3 3 0.889581 339 3 3 4 2 1 1 0.479881 672 3 2 4 1 1 4 0.64927
7 1 3 2 4 4 3 0.876967 340 3 3 4 1 3 2 0.924292 673 3 2 4 3 3 3 0.924116
8 2 3 2 1 2 4 0.625871 341 3 4 2 3 4 3 0.994732 674 3 3 2 1 3 2 0.924274
9 2 4 4 4 4 3 0.975033 342 4 3 2 3 3 3 0.988012 675 4 3 4 4 2 2 0.932044
10 1 4 2 2 3 3 0.595864 343 4 2 3 3 3 4 0.986963 676 3 3 2 1 3 2 0.924274
11 3 3 2 4 2 4 0.908474 344 4 2 3 2 3 3 0.939927 677 4 3 4 4 1 4 0.920569
12 2 3 2 2 3 3 0.889581 345 4 3 4 2 2 3 0.974601 678 3 4 3 1 3 1 0.88867
13 3 3 2 1 3 2 0.924274 346 3 4 2 2 3 3 0.978535 679 3 3 2 4 2 4 0.908474
14 1 2 3 3 3 3 0.565197 347 2 4 4 4 4 3 0.975033 680 3 4 2 1 3 2 0.965121
15 1 2 3 3 3 3 0.565197 348 3 4 2 3 3 4 0.979558 681 3 3 2 4 3 4 0.978393
16 3 3 2 4 2 4 0.908474 349 3 4 3 4 1 4 0.655498 682 3 3 4 2 1 1 0.479881
17 4 3 4 4 2 2 0.932044 350 2 4 4 2 2 3 0.625998 683 2 2 4 4 3 4 0.889102
18 4 3 4 2 3 3 0.988141 351 2 1 3 1 2 4 0.587524 684 1 4 2 3 3 3 0.598298
19 2 3 2 1 3 2 0.823504 352 4 2 3 4 1 4 0.908343 685 1 4 2 3 3 3 0.598298
20 3 3 4 2 3 3 0.971365 353 3 4 3 1 3 2 0.965124 686 2 1 3 2 3 3 0.629032
21 3 3 2 4 1 4 0.653213 354 4 3 4 2 2 3 0.974601 687 3 3 4 2 3 3 0.971365
22 3 4 2 3 4 3 0.994732 355 2 3 2 1 2 4 0.625871 688 3 3 4 1 3 2 0.924292
23 4 3 4 3 3 4 0.995332 356 3 3 2 1 3 2 0.924274 689 3 3 2 1 3 2 0.924274
24 3 3 2 3 3 4 0.978392 357 2 2 4 4 2 2 0.448921 690 1 2 3 4 4 3 0.834297
25 4 3 4 1 3 2 0.944312 358 3 3 2 4 2 4 0.908474 691 2 4 4 3 3 3 0.89658
26 3 3 2 3 3 4 0.978392 359 3 4 3 1 3 2 0.965124 692 2 1 3 2 3 3 0.629032
27 3 4 2 3 4 3 0.994732 360 3 3 2 4 2 4 0.908474 693 3 4 2 3 4 3 0.994732
28 3 3 2 1 3 2 0.924274 361 2 3 4 2 3 3 0.88959 694 4 3 4 3 4 3 0.990911
29 4 2 4 3 1 4 0.912525 362 3 4 2 3 4 3 0.994732 695 4 3 4 4 2 4 0.981685
30 3 3 2 2 1 1 0.479881 363 4 3 4 4 1 4 0.920569 696 2 3 2 1 3 2 0.823504
31 2 3 3 2 3 3 0.889589 364 4 3 4 2 3 3 0.988141 697 3 3 4 2 1 1 0.479881
32 3 3 2 1 1 4 0.653213 365 2 4 4 4 4 3 0.975033 698 3 3 4 1 3 1 0.69553
33 1 4 2 4 3 4 0.599148 366 3 3 4 1 3 1 0.69553 699 1 3 2 3 4 3 0.876408
34 1 3 2 3 3 3 0.593981 367 4 2 4 2 1 2 0.660132 700 3 4 3 4 1 4 0.655498
35 4 3 4 2 2 3 0.974601 368 2 3 2 3 3 4 0.896479 701 3 4 3 2 3 3 0.978542
36 1 2 3 3 3 3 0.565197 369 3 2 4 4 2 2 0.651811 702 4 3 4 4 4 3 0.990912
37 4 3 4 3 4 3 0.990911 370 4 3 4 1 1 4 0.920569 703 4 3 4 4 2 4 0.981685
38 3 3 2 2 1 2 0.621323 371 3 3 2 3 3 4 0.978392 704 4 3 4 4 1 4 0.920569
39 3 4 2 4 1 4 0.655348 372 4 3 4 1 3 2 0.944312 705 2 3 4 4 2 2 0.59435
40 3 4 2 2 3 3 0.978535 373 3 3 2 4 2 2 0.862879 706 4 2 3 3 3 3 0.939928
41 2 4 4 4 4 3 0.975033 374 4 2 4 1 2 4 0.973503 707 3 3 2 1 1 4 0.653213
42 3 3 2 3 4 3 0.987393 375 2 3 2 2 3 3 0.889581 708 3 4 2 3 4 3 0.994732
43 3 3 2 3 1 4 0.653213 376 4 2 3 2 2 3 0.926572 709 2 3 4 3 3 2 0.850899
44 3 3 4 3 3 4 0.978494 377 4 3 4 4 1 4 0.920569 710 3 4 3 1 3 1 0.88867
45 3 3 2 4 2 2 0.862879 378 4 3 4 2 1 2 0.874052 711 3 3 4 3 3 3 0.971428
46 4 3 4 2 2 3 0.974601 379 3 2 4 1 1 4 0.64927 712 3 3 2 4 2 4 0.908474
47 2 3 4 3 4 3 0.967772 380 3 3 2 3 3 4 0.978392 713 4 2 4 1 3 2 0.713678
48 3 4 2 3 3 4 0.979558 381 3 3 2 2 1 1 0.479881 714 4 3 4 1 1 4 0.920569
49 2 1 3 2 3 3 0.629032 382 1 3 2 3 3 3 0.593981 715 3 4 2 4 1 4 0.655348
50 3 4 2 3 4 1 0.948115 383 2 3 4 2 3 3 0.88959 716 4 3 4 4 2 4 0.981685
51 2 4 4 3 3 4 0.897479 384 4 3 4 1 3 2 0.944312 717 2 1 3 2 3 3 0.629032
52 3 3 2 2 1 1 0.479881 385 3 4 2 3 4 3 0.994732 718 3 3 2 4 2 4 0.908474
53 1 4 2 2 3 3 0.595864 386 1 2 3 3 3 3 0.565197 719 4 3 4 1 3 2 0.944312
54 1 4 2 3 3 4 0.599115 387 3 3 2 4 3 4 0.978393 720 2 3 4 2 3 3 0.88959
55 3 3 4 2 1 1 0.479881 388 3 3 4 1 3 2 0.924292 721 4 2 4 2 1 2 0.660132
56 2 3 2 1 2 4 0.625871 389 2 3 4 3 3 3 0.889995 722 1 3 2 3 3 3 0.593981
57 3 3 4 4 2 2 0.862971 390 4 3 4 4 1 4 0.920569 723 1 4 2 3 3 3 0.598298
58 2 3 4 3 3 4 0.896498 391 3 3 2 1 3 2 0.924274 724 3 2 4 4 3 4 0.970411
59 2 1 3 4 2 4 0.587524 392 3 3 2 4 3 4 0.978393 725 2 3 4 2 3 3 0.88959
60 4 3 4 1 3 2 0.944312 393 2 2 4 4 3 4 0.889102 726 3 4 2 3 4 3 0.994732
61 4 3 4 1 3 1 0.726208 394 1 4 2 3 3 4 0.599115 727 3 3 2 4 2 2 0.862879
62 1 2 3 3 4 3 0.834106 395 4 3 4 2 3 3 0.988141 728 2 2 4 3 3 3 0.846678
63 3 3 2 4 2 4 0.908474 396 3 3 2 3 2 4 0.908474 729 3 3 2 3 3 3 0.971357
64 4 3 4 2 2 3 0.974601 397 2 3 2 3 3 3 0.889986 730 3 2 4 1 1 4 0.64927
65 3 3 4 1 3 2 0.924292 398 3 3 2 4 2 4 0.908474 731 2 1 3 2 3 3 0.629032
66 3 3 4 4 1 4 0.654782 399 3 4 3 4 1 4 0.655498 732 4 2 4 2 3 2 0.71376

(continued on next page)

5
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Table A1 (continued)

Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability
1 2 1 3 2 3 3 0.629032 334 2 2 4 4 4 3 0.920716 667 3 4 2 1 1 4 0.655346
67 2 4 4 4 2 2 0.622084 400 2 3 4 1 3 2 0.823504 733 3 3 2 3 3 3 0.971357
68 2 3 4 2 2 3 0.62144 401 4 3 4 1 2 4 0.98168 734 1 2 3 3 3 3 0.565197
69 3 4 3 1 3 1 0.88867 402 3 2 1 1 3 2 0.699533 735 3 3 2 1 1 4 0.653213
70 4 2 3 2 3 3 0.939927 403 3 3 2 4 4 3 0.987403 736 2 4 4 1 2 4 0.626554
71 3 3 2 1 3 1 0.69553 404 2 3 2 1 2 4 0.625871 737 3 3 4 2 1 1 0.479881
72 4 2 3 3 4 3 0.942634 405 2 1 3 3 3 4 0.84183 738 3 2 4 1 1 4 0.64927
73 3 3 2 4 2 2 0.862879 406 2 3 2 3 3 3 0.889986 739 4 3 4 1 1 1 0.674702
74 3 3 2 4 3 4 0.978393 407 1 2 3 2 3 3 0.564669 740 3 4 3 4 2 2 0.903244
75 3 2 2 1 3 2 0.700876 408 3 3 2 1 1 2 0.621323 741 3 4 2 4 2 2 0.903232
76 4 3 4 4 1 4 0.920569 409 4 2 4 1 1 4 0.912525 742 2 3 4 2 3 3 0.88959
77 3 3 2 4 3 4 0.978393 410 3 2 4 1 1 4 0.64927 743 2 4 4 2 3 3 0.896012
78 2 2 4 2 2 3 0.591223 411 4 2 4 2 1 3 0.869154 744 3 2 4 1 3 2 0.700961
79 1 3 2 3 3 3 0.593981 412 3 2 4 4 2 4 0.901403 745 3 3 4 4 3 4 0.978495
80 3 2 4 4 1 4 0.64927 413 3 3 2 4 2 4 0.908474 746 2 3 4 3 4 3 0.967772
81 3 3 2 1 3 1 0.69553 414 3 3 2 4 2 2 0.862879 747 3 3 2 1 3 2 0.924274
82 3 3 2 2 1 2 0.621323 415 3 2 4 1 1 4 0.64927 748 3 2 4 2 1 2 0.469555
83 3 3 2 1 3 2 0.924274 416 2 3 2 1 2 4 0.625871 749 2 3 2 1 2 4 0.625871
84 2 2 4 4 3 4 0.889102 417 2 1 3 2 3 3 0.629032 750 2 4 4 4 3 4 0.897486
85 3 4 2 1 3 4 0.979428 418 2 4 4 2 3 3 0.896012 751 2 2 4 4 3 4 0.889102
86 2 1 3 2 3 3 0.629032 419 3 3 2 4 2 4 0.908474 752 2 3 2 3 3 3 0.889986
87 2 2 4 1 2 4 0.620841 420 4 3 4 2 3 3 0.988141 753 3 4 2 1 1 4 0.655346
88 2 3 4 3 4 3 0.967772 421 3 3 2 1 3 2 0.924274 754 4 3 4 2 3 3 0.988141
89 3 3 2 1 1 4 0.653213 422 3 4 2 3 3 3 0.978631 755 4 3 4 3 3 4 0.995332
90 3 3 2 4 4 2 0.944284 423 1 4 2 2 3 3 0.595864 756 3 3 2 1 3 2 0.924274
91 2 3 2 4 4 3 0.967847 424 3 2 4 3 3 3 0.924116 757 1 3 2 3 3 3 0.593981
92 1 2 3 3 3 3 0.565197 425 3 4 2 3 4 3 0.994732 758 3 3 2 4 2 2 0.862879
93 3 4 3 4 4 3 0.994748 426 2 2 4 4 3 4 0.889102 759 3 3 2 2 3 3 0.971294
94 2 1 3 2 3 3 0.629032 427 4 2 4 3 3 4 0.987103 760 4 3 4 2 3 3 0.988141
95 3 3 2 1 3 1 0.69553 428 3 4 3 3 1 4 0.655498 761 2 3 2 4 2 2 0.59435
96 2 2 4 3 3 3 0.846678 429 3 3 2 1 3 2 0.924274 762 4 2 4 3 1 4 0.912525
97 4 2 3 2 1 2 0.659402 430 3 2 4 3 3 4 0.970411 763 3 3 2 4 2 2 0.862879
98 2 3 4 4 3 4 0.896503 431 3 2 2 4 1 4 0.635641 764 2 4 4 3 4 3 0.974923
99 3 4 3 1 3 1 0.88867 432 3 2 4 3 4 3 0.939321 765 3 4 3 1 1 4 0.655496
100 3 3 1 4 2 2 0.861433 433 3 4 2 3 4 3 0.994732 766 3 3 2 4 2 4 0.908474
101 2 2 4 3 3 4 0.889101 434 2 1 3 4 3 4 0.84183 767 3 3 2 2 3 3 0.971294
102 1 4 2 3 3 3 0.598298 435 2 2 4 3 3 3 0.846678 768 2 3 4 1 3 2 0.823504
103 4 3 4 1 3 1 0.726208 436 4 3 4 3 3 4 0.995332 769 2 1 3 2 3 3 0.629032
104 4 2 4 2 2 3 0.92709 437 3 2 4 1 1 4 0.64927 770 2 3 4 2 3 3 0.88959
105 3 3 2 4 2 4 0.908474 438 1 2 3 3 3 3 0.565197 771 3 4 3 1 3 2 0.965124
106 4 2 4 1 1 4 0.912525 439 1 2 3 4 3 4 0.593553 772 3 3 2 2 3 3 0.971294
107 3 2 4 4 2 4 0.901403 440 3 3 4 1 2 4 0.908911 773 3 4 2 1 3 2 0.965121
108 2 2 4 4 3 4 0.889102 441 4 3 4 1 1 4 0.920569 774 1 2 3 3 3 3 0.565197
109 1 3 2 2 3 3 0.592089 442 2 3 2 1 2 4 0.625871 775 4 3 4 3 3 3 0.98815
110 3 3 2 4 3 4 0.978393 443 2 4 4 1 3 2 0.853197 776 2 3 4 2 3 3 0.88959
111 4 3 4 3 3 4 0.995332 444 4 2 3 2 2 3 0.926572 777 3 3 2 1 3 1 0.69553
112 2 3 4 2 3 3 0.88959 445 1 2 3 3 3 3 0.565197 778 3 4 3 1 3 1 0.88867
113 3 3 2 1 1 4 0.653213 446 4 2 3 3 1 4 0.908343 779 4 3 4 4 2 2 0.932044
114 1 3 2 4 3 4 0.598492 447 2 3 3 2 3 3 0.889589 780 3 3 2 3 3 3 0.971357
115 3 3 2 2 3 3 0.971294 448 3 4 3 1 3 2 0.965124 781 3 4 3 4 2 2 0.903244
116 2 2 4 2 2 3 0.591223 449 3 2 4 2 3 3 0.924103 782 3 4 2 2 3 3 0.978535
117 3 2 4 1 1 4 0.64927 450 4 3 4 4 2 2 0.932044 783 3 4 3 1 3 1 0.88867
118 3 3 2 3 1 4 0.653213 451 2 1 3 2 3 3 0.629032 784 3 3 2 2 3 3 0.971294
119 2 4 4 2 2 3 0.625998 452 2 1 3 2 3 3 0.629032 785 3 4 3 1 2 4 0.909892
120 3 3 2 4 2 4 0.908474 453 4 2 3 2 3 4 0.986962 786 3 2 4 3 4 3 0.939321
121 3 3 2 4 2 2 0.862879 454 3 4 3 3 3 4 0.979567 787 3 4 3 1 3 1 0.88867
122 3 2 4 3 4 3 0.939321 455 1 3 2 2 3 3 0.592089 788 4 3 4 1 3 1 0.726208
123 3 4 3 3 1 4 0.655498 456 2 2 4 2 3 3 0.846583 789 3 3 2 4 2 4 0.908474
124 3 2 4 1 1 4 0.64927 457 3 3 2 2 3 3 0.971294 790 3 3 2 4 2 2 0.862879
125 2 2 4 1 3 2 0.637487 458 2 1 3 4 2 4 0.587524 791 3 4 1 1 1 4 0.653701
126 3 3 3 4 2 4 0.908899 459 2 3 4 3 3 4 0.896498 792 2 4 4 2 2 3 0.625998
127 2 1 3 2 3 3 0.629032 460 1 3 2 3 4 3 0.876408 793 3 4 3 1 1 4 0.655496
128 4 2 3 4 2 4 0.972714 461 4 2 4 4 3 4 0.987103 794 2 1 3 2 3 3 0.629032
129 3 2 4 4 2 2 0.651811 462 4 2 3 4 3 4 0.986963 795 2 3 4 2 3 3 0.88959
130 4 3 4 4 1 4 0.920569 463 2 4 4 4 4 3 0.975033 796 3 3 2 2 3 1 0.714909
131 1 4 2 3 3 4 0.599115 464 3 3 2 1 3 1 0.69553 797 3 3 4 2 1 1 0.479881
132 4 2 3 2 2 3 0.926572 465 4 2 4 4 1 4 0.912525 798 3 4 2 3 4 3 0.994732
133 3 3 2 4 2 2 0.862879 466 3 4 2 4 4 3 0.994747 799 4 3 4 3 4 3 0.990911
134 4 3 4 3 3 3 0.98815 467 3 3 4 4 2 4 0.908937 800 4 3 4 1 3 1 0.726208
135 3 3 2 4 2 2 0.862879 468 4 3 4 1 1 4 0.920569 801 4 2 3 1 1 4 0.908343
136 4 3 4 3 3 4 0.995332 469 3 2 4 2 3 3 0.924103 802 3 3 2 4 4 3 0.987403
137 2 1 3 4 3 4 0.84183 470 2 3 2 2 2 3 0.621416 803 3 4 2 4 1 4 0.655348
138 3 3 2 4 1 4 0.653213 471 2 3 4 2 3 3 0.88959 804 2 4 4 4 2 4 0.626698
139 2 1 3 2 3 3 0.629032 472 2 1 3 2 3 3 0.629032 805 4 2 3 1 2 4 0.972714
140 3 2 4 4 4 3 0.939324 473 3 3 2 1 3 2 0.924274 806 2 4 4 4 4 3 0.975033

6
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Table A1 (continued)

Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability
1 2 1 3 2 3 3 0.629032 334 2 2 4 4 4 3 0.920716 667 3 4 2 1 1 4 0.655346
141 3 2 4 3 1 4 0.64927 474 3 3 2 1 3 1 0.69553 807 3 4 2 2 1 1 0.624578
142 1 3 2 2 4 3 0.871486 475 2 1 2 1 2 4 0.582446 808 4 2 3 3 3 3 0.939928
143 2 1 3 2 3 3 0.629032 476 2 1 3 2 3 3 0.629032 809 1 4 2 3 4 3 0.88272
144 3 4 3 4 2 3 0.90908 477 4 2 3 4 4 3 0.942634 810 3 3 2 2 2 3 0.902026
145 3 4 3 1 3 1 0.88867 478 3 2 4 1 3 2 0.700961 811 4 3 4 1 1 4 0.920569
146 4 2 4 2 3 3 0.940014 479 4 2 4 1 3 2 0.713678 812 1 2 3 4 4 3 0.834297
147 4 2 3 3 3 4 0.986963 480 2 3 2 1 3 2 0.823504 813 1 2 3 3 3 3 0.565197
148 4 3 4 4 2 4 0.981685 481 3 3 2 3 4 3 0.987393 814 3 4 3 4 2 4 0.909933
149 3 3 2 1 3 1 0.69553 482 1 4 2 4 4 3 0.883451 815 3 3 2 1 3 2 0.924274
150 3 4 3 4 3 4 0.979568 483 4 2 4 2 3 3 0.940014 816 4 2 3 3 4 3 0.942634
151 3 3 2 2 1 2 0.621323 484 1 4 2 4 3 4 0.599148 817 3 3 2 1 1 4 0.653213
152 4 2 3 3 3 3 0.939928 485 3 4 2 4 4 3 0.994747 818 1 4 2 3 3 4 0.599115
153 3 3 2 3 3 3 0.971357 486 4 3 4 1 3 2 0.944312 819 4 3 4 2 2 3 0.974601
154 3 4 3 1 1 4 0.655496 487 3 3 4 4 4 3 0.987416 820 2 3 4 1 2 4 0.625918
155 2 2 4 4 3 4 0.889102 488 2 3 4 3 3 3 0.889995 821 3 3 2 2 1 2 0.621323
156 2 1 3 2 3 3 0.629032 489 1 4 2 4 3 4 0.599148 822 2 2 4 4 3 4 0.889102
157 2 3 4 1 3 2 0.823504 490 1 2 3 3 3 3 0.565197 823 3 3 2 2 1 1 0.479881
158 2 3 2 3 3 3 0.889986 491 1 4 2 3 3 4 0.599115 824 3 2 4 4 4 3 0.939324
159 3 3 2 4 2 2 0.862879 492 3 3 2 4 3 4 0.978393 825 2 2 4 3 3 3 0.846678
160 3 3 2 2 1 2 0.621323 493 4 2 4 3 1 4 0.912525 826 3 4 3 1 3 2 0.965124
161 4 3 4 4 1 4 0.920569 494 4 3 4 2 1 2 0.874052 827 2 4 4 2 3 3 0.896012
162 1 3 2 4 3 4 0.598492 495 4 2 4 3 3 4 0.987103 828 2 2 4 2 3 3 0.846583
163 2 1 3 2 3 3 0.629032 496 2 3 2 4 3 4 0.896484 829 3 3 4 2 1 1 0.479881
164 3 3 2 1 3 4 0.97829 497 4 3 4 2 1 1 0.674702 830 2 1 3 2 3 3 0.629032
165 2 4 4 3 3 4 0.897479 498 4 3 4 4 4 3 0.990912 831 4 3 4 1 2 4 0.98168
166 2 1 3 1 2 4 0.587524 499 4 2 4 3 3 4 0.987103 832 2 3 4 1 2 4 0.625918
167 3 4 3 4 1 4 0.655498 500 3 3 4 1 3 2 0.924292 833 4 2 4 3 1 4 0.912525
168 4 2 3 1 3 2 0.713662 501 4 3 4 4 1 4 0.920569 834 4 3 4 2 1 1 0.674702
169 3 4 3 2 2 3 0.909048 502 4 2 4 3 4 3 0.942648 835 2 1 3 4 4 3 0.684064
170 4 3 4 1 2 4 0.98168 503 3 3 2 4 4 3 0.987403 836 3 3 4 4 4 3 0.987416
171 3 2 4 3 3 4 0.970411 504 3 2 4 3 4 3 0.939321 837 3 3 4 1 2 4 0.908911
172 2 3 2 1 2 4 0.625871 505 3 4 2 4 1 4 0.655348 838 3 2 4 3 3 3 0.924116
173 3 3 2 1 3 2 0.924274 506 3 3 2 1 3 2 0.924274 839 2 2 4 2 3 3 0.846583
174 3 2 4 4 3 4 0.970411 507 3 4 2 2 1 1 0.624578 840 2 1 3 2 3 3 0.629032
175 2 3 1 4 4 3 0.96781 508 3 2 4 1 3 2 0.700961 841 3 3 4 2 1 1 0.479881
176 4 2 3 3 3 3 0.939928 509 3 2 4 3 3 4 0.970411 842 2 4 4 1 3 2 0.853197
177 3 3 4 1 3 2 0.924292 510 3 3 2 4 2 2 0.862879 843 1 4 2 3 3 4 0.599115
178 2 4 4 3 3 3 0.89658 511 2 3 2 4 4 3 0.967847 844 3 3 2 2 1 1 0.479881
179 2 2 4 3 3 4 0.889101 512 3 3 2 1 1 4 0.653213 845 2 4 4 2 3 3 0.896012
180 1 4 3 3 3 4 0.599115 513 3 4 2 4 2 4 0.90989 846 3 3 4 2 3 3 0.971365
181 3 4 2 1 3 2 0.965121 514 3 3 2 3 3 4 0.978392 847 3 4 2 3 3 3 0.978631
182 3 4 2 3 3 4 0.979558 515 1 2 3 4 4 3 0.834297 848 2 1 3 2 3 3 0.629032
183 3 3 4 4 2 4 0.908937 516 3 3 2 3 3 3 0.971357 849 3 3 2 4 3 4 0.978393
184 3 3 3 4 2 2 0.862965 517 3 3 2 4 4 3 0.987403 850 4 3 4 1 3 2 0.944312
185 4 3 4 1 3 1 0.726208 518 3 4 3 4 2 2 0.903244 851 4 2 3 2 2 3 0.926572
186 4 2 4 4 2 2 0.703977 519 2 3 2 1 3 2 0.823504 852 1 3 2 2 3 3 0.592089
187 4 2 4 2 2 3 0.92709 520 3 3 2 4 2 4 0.908474 853 4 2 3 2 2 3 0.926572
188 3 4 2 4 1 4 0.655348 521 1 2 3 3 3 3 0.565197 854 4 3 4 2 1 1 0.674702
189 4 2 3 4 2 4 0.972714 522 1 4 2 3 4 3 0.88272 855 2 3 2 4 3 4 0.896484
190 3 3 4 1 3 1 0.69553 523 1 3 2 3 3 3 0.593981 856 2 4 4 3 4 3 0.974923
191 4 3 4 4 3 4 0.995332 524 3 2 4 2 1 2 0.469555 857 2 2 4 2 3 3 0.846583
192 3 3 2 4 2 2 0.862879 525 3 3 2 3 3 3 0.971357 858 2 1 3 4 2 4 0.587524
193 4 3 4 1 3 2 0.944312 526 4 3 4 4 1 4 0.920569 859 3 4 2 1 2 4 0.909848
194 3 3 2 1 3 2 0.924274 527 2 3 4 2 3 3 0.88959 860 3 3 4 4 4 3 0.987416
195 3 4 3 1 3 1 0.88867 528 4 3 4 4 1 4 0.920569 861 4 2 4 3 1 4 0.912525
196 3 2 4 2 2 3 0.858414 529 4 3 4 3 3 3 0.98815 862 4 2 3 2 2 3 0.926572
197 3 2 4 3 4 3 0.939321 530 4 3 4 3 4 3 0.990911 863 2 3 4 3 3 3 0.889995
198 3 3 2 3 4 3 0.987393 531 3 4 2 3 1 3 0.654764 864 3 4 3 1 1 1 0.621705
199 3 3 2 1 3 1 0.69553 532 4 3 4 3 3 3 0.98815 865 2 1 3 3 3 3 0.629032
200 3 3 2 4 2 4 0.908474 533 1 2 3 3 3 3 0.565197 866 4 2 4 4 2 2 0.703977
201 4 3 4 2 3 3 0.988141 534 4 3 4 3 3 3 0.98815 867 3 3 2 4 2 4 0.908474
202 2 2 4 2 3 3 0.846583 535 4 3 4 4 1 4 0.920569 868 4 2 4 3 1 4 0.912525
203 2 3 4 4 2 2 0.59435 536 4 3 4 1 3 2 0.944312 869 4 3 4 3 1 4 0.920569
204 1 3 2 4 3 4 0.598492 537 1 3 2 3 4 3 0.876408 870 3 3 2 3 1 4 0.653213
205 3 3 4 1 1 1 0.479881 538 3 4 2 4 2 4 0.90989 871 4 2 3 1 1 4 0.908343
206 3 3 2 3 3 3 0.971357 539 4 3 4 4 1 4 0.920569 872 4 3 4 3 1 4 0.920569
207 4 2 4 2 3 3 0.940014 540 1 2 3 3 3 3 0.565197 873 1 4 2 3 3 4 0.599115
208 3 4 2 3 4 3 0.994732 541 1 4 2 3 3 4 0.599115 874 3 3 2 3 3 4 0.978392
209 3 1 2 3 3 3 0.683162 542 3 3 2 2 2 3 0.902026 875 2 1 3 2 3 3 0.629032
210 3 3 4 4 4 3 0.987416 543 1 3 2 3 3 3 0.593981 876 3 3 2 4 2 2 0.862879
211 3 3 4 3 3 4 0.978494 544 4 3 4 1 1 4 0.920569 877 2 3 4 2 3 3 0.88959
212 2 4 4 4 4 3 0.975033 545 3 3 2 2 1 2 0.621323 878 4 3 4 1 3 2 0.944312
213 4 3 4 2 3 3 0.988141 546 4 2 4 3 1 4 0.912525 879 3 3 2 2 3 3 0.971294
214 3 3 2 1 1 4 0.653213 547 3 3 2 4 1 2 0.621323 880 3 3 2 1 3 2 0.924274

(continued on next page)

7
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Table A1 (continued)

Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability
1 2 1 3 2 3 3 0.629032 334 2 2 4 4 4 3 0.920716 667 3 4 2 1 1 4 0.655346
215 2 3 4 2 3 2 0.848303 548 2 3 4 2 3 3 0.88959 881 3 2 4 1 1 4 0.64927
216 4 2 4 2 3 3 0.940014 549 3 2 4 1 1 4 0.64927 882 1 3 2 3 3 3 0.593981
217 3 3 2 2 3 3 0.971294 550 4 3 4 1 3 2 0.944312 883 3 3 4 3 1 4 0.654782
218 4 3 4 4 1 4 0.920569 551 4 2 3 2 3 3 0.939927 884 3 3 4 2 1 1 0.479881
219 3 3 2 4 4 3 0.987403 552 4 3 4 4 2 2 0.932044 885 1 3 2 3 3 3 0.593981
220 1 2 3 3 3 3 0.565197 553 3 3 2 4 2 2 0.862879 886 3 3 2 3 1 4 0.653213
221 3 3 2 1 3 2 0.924274 554 3 3 2 4 4 3 0.987403 887 3 3 4 4 3 4 0.978495
222 1 2 3 4 4 3 0.834297 555 2 3 2 1 2 4 0.625871 888 2 2 4 4 3 4 0.889102
223 4 3 4 3 4 3 0.990911 556 2 2 4 4 2 2 0.448921 889 2 2 4 3 4 2 0.699025
224 3 3 2 3 1 4 0.653213 557 3 3 2 1 3 2 0.924274 890 3 3 4 4 3 4 0.978495
225 3 2 4 2 1 2 0.469555 558 4 2 4 4 1 4 0.912525 891 2 2 4 2 2 3 0.591223
226 3 3 2 1 3 1 0.69553 559 2 3 2 3 3 3 0.889986 892 3 3 4 2 3 3 0.971365
227 3 3 2 2 1 2 0.621323 560 3 4 3 4 1 4 0.655498 893 2 2 4 1 2 4 0.620841
228 1 4 2 3 3 3 0.598298 561 3 3 2 2 1 2 0.621323 894 2 4 4 1 3 2 0.853197
229 2 3 2 3 4 3 0.96777 562 3 3 2 3 3 3 0.971357 895 3 3 2 2 1 2 0.621323
230 4 2 3 2 1 2 0.659402 563 3 3 2 2 2 3 0.902026 896 2 4 4 2 3 3 0.896012
231 3 4 3 3 3 4 0.979567 564 3 3 2 1 1 4 0.653213 897 3 3 2 1 1 4 0.653213
232 2 2 4 3 3 4 0.889101 565 4 2 4 3 1 4 0.912525 898 3 3 2 2 1 1 0.479881
233 3 2 4 4 4 2 0.713236 566 3 4 3 4 1 2 0.65069 899 4 2 4 3 1 4 0.912525
234 2 3 2 4 4 3 0.967847 567 1 3 2 3 3 3 0.593981 900 4 2 3 3 3 3 0.939928
235 1 2 3 4 4 3 0.834297 568 4 2 3 2 1 2 0.659402 901 2 1 3 2 3 3 0.629032
236 4 2 3 3 3 4 0.986963 569 4 2 3 2 2 3 0.926572 902 4 3 4 1 3 2 0.944312
237 3 3 2 4 3 4 0.978393 570 3 3 2 3 4 3 0.987393 903 2 1 3 4 4 3 0.684064
238 3 4 3 4 2 4 0.909933 571 1 2 3 4 3 4 0.593553 904 3 3 2 3 3 4 0.978392
239 3 4 2 2 1 2 0.650624 572 1 2 3 4 3 2 0.429178 905 3 3 2 2 3 3 0.971294
240 3 2 4 1 1 4 0.64927 573 3 3 4 2 1 1 0.479881 906 3 3 2 1 3 2 0.924274
241 3 4 3 3 3 3 0.978638 574 3 3 2 1 3 2 0.924274 907 3 3 2 2 1 2 0.621323
242 4 3 4 1 3 2 0.944312 575 3 2 4 3 4 3 0.939321 908 3 3 2 3 3 4 0.978392
243 4 2 4 4 4 3 0.942648 576 1 3 2 4 3 4 0.598492 909 4 2 3 3 3 4 0.986963
244 3 3 2 3 3 3 0.971357 577 3 4 3 1 3 1 0.88867 910 3 3 2 2 1 2 0.621323
245 1 2 3 3 3 3 0.565197 578 3 3 2 1 1 4 0.653213 911 2 2 4 4 3 4 0.889102
246 2 3 2 3 3 3 0.889986 579 4 2 3 2 1 2 0.659402 912 4 2 4 4 4 3 0.942648
247 2 2 4 4 2 4 0.620841 580 3 4 2 1 3 1 0.88867 913 3 4 3 1 1 4 0.655496
248 4 2 4 4 2 4 0.973503 581 3 2 4 3 3 3 0.924116 914 3 4 2 2 2 3 0.909013
249 1 3 2 3 3 3 0.593981 582 2 2 3 3 3 3 0.846673 915 3 2 4 1 1 4 0.64927
250 2 1 3 4 3 4 0.84183 583 3 4 3 2 3 3 0.978542 916 2 2 4 4 2 2 0.448921
251 3 4 3 4 4 3 0.994748 584 2 1 3 2 3 3 0.629032 917 2 1 3 4 3 4 0.84183
252 3 4 3 3 3 3 0.978638 585 1 3 2 3 3 3 0.593981 918 3 3 2 2 2 4 0.908472
253 3 3 2 3 1 4 0.653213 586 3 4 1 1 1 4 0.653701 919 3 3 2 4 2 2 0.862879
254 2 4 4 4 2 4 0.626698 587 3 3 2 4 2 2 0.862879 920 2 2 4 1 3 2 0.637487
255 2 1 3 2 3 3 0.629032 588 4 3 4 4 4 3 0.990912 921 3 4 2 1 3 2 0.965121
256 4 3 4 1 3 1 0.726208 589 3 4 3 3 1 4 0.655498 922 3 3 2 3 3 3 0.971357
257 2 4 4 4 4 3 0.975033 590 3 3 2 2 3 3 0.971294 923 4 3 4 4 2 2 0.932044
258 4 3 4 3 3 4 0.995332 591 1 3 2 2 3 3 0.592089 924 3 3 4 4 2 2 0.862971
259 2 1 3 3 3 3 0.629032 592 2 1 3 2 2 3 0.439172 925 3 3 2 4 4 3 0.987403
260 3 4 2 2 1 1 0.624578 593 1 2 3 2 3 3 0.564669 926 2 1 3 4 2 4 0.587524
261 3 3 2 4 3 4 0.978393 594 2 4 4 2 2 3 0.625998 927 3 4 2 3 4 3 0.994732
262 1 3 2 4 3 4 0.598492 595 4 2 4 3 3 4 0.987103 928 4 2 4 2 1 2 0.660132
263 3 3 2 2 1 1 0.479881 596 4 2 3 4 2 2 0.703866 929 3 3 2 4 2 2 0.862879
264 1 3 2 3 3 3 0.593981 597 4 2 4 2 3 3 0.940014 930 1 4 2 3 3 4 0.599115
265 4 3 4 4 3 4 0.995332 598 1 2 3 3 3 3 0.565197 931 3 2 4 4 2 4 0.901403
266 1 2 2 4 3 4 0.593546 599 1 2 3 3 3 4 0.593547 932 1 2 3 2 3 3 0.564669
267 3 3 4 3 1 4 0.654782 600 4 3 4 1 3 2 0.944312 933 4 3 4 2 3 3 0.988141
268 3 4 3 3 1 4 0.655498 601 3 3 2 3 3 3 0.971357 934 1 3 2 3 3 3 0.593981
269 4 3 4 4 4 3 0.990912 602 4 2 4 4 4 3 0.942648 935 2 1 3 3 3 3 0.629032
270 3 3 2 1 3 2 0.924274 603 4 3 4 2 1 1 0.674702 936 2 3 4 3 3 3 0.889995
271 3 3 2 4 2 2 0.862879 604 4 3 4 4 2 2 0.932044 937 2 1 3 2 2 3 0.439172
272 3 3 2 3 1 4 0.653213 605 2 3 4 1 2 4 0.625918 938 1 2 3 3 3 3 0.565197
273 4 3 4 4 2 4 0.981685 606 4 2 4 1 2 4 0.973503 939 1 3 2 3 3 3 0.593981
274 2 4 4 4 3 4 0.897486 607 4 2 3 4 2 2 0.703866 940 1 2 3 3 3 3 0.565197
275 4 3 4 2 1 1 0.674702 608 3 3 2 1 3 2 0.924274 941 1 3 2 2 3 3 0.592089
276 2 4 4 1 3 2 0.853197 609 2 4 4 4 4 3 0.975033 942 3 4 3 1 1 4 0.655496
277 3 3 4 3 3 3 0.971428 610 3 3 2 4 2 2 0.862879 943 2 3 4 4 4 3 0.96785
278 3 3 2 3 3 3 0.971357 611 3 2 4 2 1 2 0.469555 944 3 2 4 1 1 4 0.64927
279 1 4 2 3 3 4 0.599115 612 3 3 2 4 1 4 0.653213 945 3 1 4 3 4 3 0.697894
280 2 4 4 4 4 3 0.975033 613 4 3 4 1 3 2 0.944312 946 4 3 4 1 3 1 0.726208
281 2 1 3 1 2 4 0.587524 614 3 3 2 1 3 2 0.924274 947 3 3 2 1 1 4 0.653213
282 2 3 2 2 2 3 0.621416 615 3 3 1 1 3 2 0.923992 948 1 3 2 3 3 3 0.593981
283 4 2 4 3 3 4 0.987103 616 4 3 4 4 2 4 0.981685 949 4 2 4 1 3 2 0.713678
284 3 3 4 2 3 3 0.971365 617 3 3 4 1 3 2 0.924292 950 4 3 4 1 3 2 0.944312
285 2 3 4 2 3 3 0.88959 618 3 3 3 4 3 4 0.978487 951 2 3 4 1 3 2 0.823504
286 4 2 3 1 1 4 0.908343 619 3 3 2 4 3 4 0.978393 952 3 3 2 1 3 2 0.924274
287 3 3 2 2 1 1 0.479881 620 3 3 4 1 3 2 0.924292 953 3 2 4 2 3 3 0.924103
288 3 3 4 4 2 2 0.862971 621 3 3 2 2 1 2 0.621323 954 1 3 2 3 3 3 0.593981

8
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

Table A1 (continued)

Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability Data No. Maximum capacities Exact Reliability
1 2 1 3 2 3 3 0.629032 334 2 2 4 4 4 3 0.920716 667 3 4 2 1 1 4 0.655346
289 3 3 4 2 2 3 0.902365 622 3 3 2 2 1 2 0.621323 955 3 4 2 3 3 4 0.979558
290 3 3 2 4 1 4 0.653213 623 2 3 2 1 2 4 0.625871 956 1 3 2 2 3 3 0.592089
291 4 2 4 4 1 4 0.912525 624 2 2 4 3 3 3 0.846678 957 3 4 3 1 3 2 0.965124
292 2 3 2 1 2 4 0.625871 625 3 3 4 3 3 4 0.978494 958 4 3 4 3 3 3 0.98815
293 3 3 2 2 3 3 0.971294 626 2 2 4 4 2 2 0.448921 959 3 4 2 4 2 4 0.90989
294 4 2 4 1 3 2 0.713678 627 3 3 2 4 2 2 0.862879 960 1 3 2 3 3 3 0.593981
295 4 3 4 1 3 2 0.944312 628 4 2 3 4 3 4 0.986963 961 1 2 3 4 4 3 0.834297
296 2 2 4 2 3 3 0.846583 629 3 4 2 1 1 4 0.655346 962 3 3 2 4 2 2 0.862879
297 2 2 4 1 3 2 0.637487 630 3 4 2 3 3 3 0.978631 963 3 3 2 2 3 3 0.971294
298 3 3 4 2 1 1 0.479881 631 4 3 4 1 2 4 0.98168 964 2 3 2 3 4 3 0.96777
299 4 2 4 3 1 4 0.912525 632 4 2 3 1 2 4 0.972714 965 4 3 4 1 3 2 0.944312
300 1 2 3 3 3 3 0.565197 633 2 1 3 2 3 3 0.629032 966 3 2 4 2 3 3 0.924103
301 3 2 4 4 1 4 0.64927 634 1 2 3 3 3 3 0.565197 967 3 3 2 4 2 2 0.862879
302 3 3 2 1 3 2 0.924274 635 2 2 4 1 2 4 0.620841 968 2 4 4 3 3 3 0.89658
303 2 3 2 3 3 3 0.889986 636 1 4 2 3 3 4 0.599115 969 2 4 4 1 3 2 0.853197
304 4 3 4 2 1 2 0.874052 637 2 2 4 1 3 2 0.637487 970 2 4 4 1 2 4 0.626554
305 2 3 4 4 3 4 0.896503 638 3 4 2 4 2 4 0.90989 971 3 4 3 2 3 3 0.978542
306 2 1 3 3 3 4 0.84183 639 2 3 4 2 3 3 0.88959 972 3 3 4 2 1 1 0.479881
307 2 4 4 4 4 3 0.975033 640 2 2 4 1 2 4 0.620841 973 1 4 2 3 4 3 0.88272
308 4 3 4 1 3 1 0.726208 641 4 3 4 3 3 3 0.98815 974 3 3 2 1 3 2 0.924274
309 3 2 4 1 3 2 0.700961 642 3 3 2 4 2 2 0.862879 975 3 4 2 1 3 2 0.965121
310 3 3 2 1 3 2 0.924274 643 4 2 3 4 3 4 0.986963 976 1 4 2 3 3 4 0.599115
311 4 3 4 4 1 4 0.920569 644 4 3 4 1 1 4 0.920569 977 4 3 4 4 1 4 0.920569
312 2 1 3 4 4 3 0.684064 645 1 4 2 3 3 4 0.599115 978 1 2 3 4 4 2 0.633439
313 2 1 3 3 4 3 0.684063 646 2 4 4 1 3 2 0.853197 979 1 4 2 4 3 4 0.599148
314 3 4 2 3 3 4 0.979558 647 4 2 3 1 1 4 0.908343 980 4 3 4 1 1 4 0.920569
315 1 2 3 2 3 3 0.564669 648 3 4 3 1 3 2 0.965124 981 3 3 2 4 4 3 0.987403
316 3 3 2 4 2 4 0.908474 649 4 2 3 1 3 2 0.713662 982 1 2 3 3 3 3 0.565197
317 3 3 2 4 2 4 0.908474 650 2 2 4 4 3 4 0.889102 983 2 4 4 4 4 3 0.975033
318 3 2 4 1 2 4 0.901403 651 3 2 4 4 4 3 0.939324 984 3 3 2 2 1 2 0.621323
319 2 2 4 4 3 4 0.889102 652 3 3 2 1 3 2 0.924274 985 1 1 3 2 3 3 0.419958
320 3 3 2 2 1 2 0.621323 653 3 3 2 3 3 3 0.971357 986 3 4 3 1 3 1 0.88867
321 3 3 2 2 3 3 0.971294 654 4 3 4 4 2 4 0.981685 987 4 3 4 1 3 2 0.944312
322 4 2 3 2 2 3 0.926572 655 3 2 4 4 3 4 0.970411 988 4 3 4 4 2 4 0.981685
323 3 3 2 1 3 2 0.924274 656 1 2 3 3 3 3 0.565197 989 3 3 2 1 3 1 0.69553
324 3 3 4 3 4 3 0.987406 657 3 3 2 1 3 2 0.924274 990 3 4 3 2 1 2 0.65067
325 2 2 4 4 3 4 0.889102 658 4 2 3 2 2 3 0.926572 991 1 2 3 3 3 3 0.565197
326 2 3 2 1 2 4 0.625871 659 3 3 2 4 2 2 0.862879 992 2 1 3 1 2 4 0.587524
327 3 3 2 4 3 4 0.978393 660 3 2 4 4 2 2 0.651811 993 1 2 3 3 3 3 0.565197
328 3 3 2 4 2 4 0.908474 661 3 3 2 2 3 3 0.971294 994 1 4 2 1 3 3 0.575998
329 3 3 2 1 1 4 0.653213 662 3 3 2 2 3 3 0.971294 995 3 3 2 1 3 2 0.924274
330 4 3 4 3 4 3 0.990911 663 3 3 3 4 1 4 0.654653 996 3 4 3 1 3 1 0.88867
331 4 2 4 1 3 2 0.713678 664 3 3 4 4 4 3 0.987416 997 4 3 4 4 1 4 0.920569
332 4 2 4 4 2 2 0.703977 665 2 4 4 4 4 3 0.975033 998 2 3 2 4 2 2 0.59435
333 2 4 4 3 4 3 0.974923 666 4 2 4 4 2 4 0.973503 999 4 3 4 2 1 2 0.874052
1000 1 4 2 4 3 4 0.599148

References [11] Shin-Guang. Chen. 2014. Optimal double-resource assignment for the robust
design problem in multistate computer networks. Applied Mathematical
Modeling 38 (1) 263–277.
[1] Lin JS, Jane CC, Yuan J. On reliability evaluation of a capacitated-flow network
[12] Lin Y-K, Yeh C-T. System reliability maximization for a computer network by
in terms of minimal pathsets. Networks 1995;25(3):131–8.
finding the optimal two-class allocation subject to budget. Appl Soft Comput
[2] Lin YK. A simple algorithm for reliability evaluation of a stochastic-flow
2015;36:578–88.
network with node failure. Comput Oper Res 2001;28(13):1277–85.
[13] Hassan MR. Solving a Component Assignment Problem for a Stochastic-
[3] Jane CC, Lin JS, Yuan J. Reliability evaluation of a limited-flow network in terms
Flow Network under Lead-time Constraint. Indian J Sci Technol 2015;8
of minimal cutsets. IEEE Trans Reliab 1993;42(3):354–61.
(35):1–5.
[4] Lin YK. On reliability evaluation of a stochastic-flow network in terms of
[14] Lin YK, Yeh CT. A Two-Stage Approach for a Multi-Objective Component
minimal cuts. J Chin Inst Industr Eng 2001;18(3):49–54.
Assignment Problem for a Stochastic-Flow Network. Eng Optim 2013;45
[5] Chu PC, Beasley JE. Genetic algorithm for generalized assignment problem.
(3):265–85.
Comput Oper Res 1997;24(1):17–23.
[15] Chen SG. An optimal capacity assignment for the robust design problem in
[6] Wang YZ. An application of genetic algorithm methods for teacher assignment
capacitated flow networks. App Math Model 2012;36(11):5272–82.
problems. Expert Syst Appl 2002;22(4):295–302.
[16] Chen, S. G., & Lin, Y. K. (2008). Capacity assignment for a stochastic-flow
[7] Harper PR, de Senna V, Vieira IT, Shahani AK. genetic algorithm for the project
network based on structural importance. In Asian International Workshop on
assignment problem. Comput Oper Res 2005;32(5):1255–65.
Advanced Reliability Modeling.
[8] Lin YK, Yeh C-T. Evaluation of optimal network reliability under components
[17] Chen SG, Lin YK. An approximate algorithm for the robust design in a
assignments subject to a transmission Budget. Transactions on Reliability
stochastic-flow network. Communications in Statistics—Theory and Methods
2010;59(3):539–50.
2010;39(13):2440–54.
[9] Lin YK, Yeh CT. Multistate Components Assignment Problem with Optimal
[18] Hamed AY, Alkinani MH, Hassan MR. A Genetic Algorithm to Solve Capacity
Network Reliability Subject to Assignment Budget. Appl Math Comput
Assignment Problem in a Flow Network. CMC-COMPUTERS MATERIALS &
2011;217(24):10074–86.
CONTINUA 2020;64(3):1579–86.
[10] Lin YK, Yeh CT. Maximizing Network Reliability for Stochastic Transportation
[19] Hamdy N, Hassan MR, Hussein ME. A genetic algorithm to solve the robust
Networks under a Budget Constraint by using a Genetic Algorithm.
design problem for a Flow Network with Node Failure. Transactions on
International Journal of Innovative Computing, Information and Control ICIC
Networks and communications 2020;8(4):1–10.
2011;7(12):7033–50.

9
M. Refaat Hassan, S. Alkhalaf, A. Mohamed Hemeida et al. Ain Shams Engineering Journal 14 (2023) 102055

[20] Hassan MR, Hamdy N, Hussein ME. Solving the robust design problem for a [34] Mona Rafat Elkharbotly. Mohamed Seddik, Abdelkawi Khalifa, ‘‘Toward
Two-Commodity Flow Network with Node Failure. American Journal of Sustainable Water: Prediction of non-revenue water via Artificial Neural
Engineering and Applied Sciences 2020.845.;13(4):837. Network and Multiple Linear Regression modelling approach in Egypt”. Ain
[21] Colbourn CJ. The Combinatorics of network Reliability,‘‘. New York: Oxford Shams Eng J 2022;13:101673.
University; 1987. [35] Pervez M, Kamal T, Fernández-Ramírez LM. novel switched model predictive
[22] Fisher ML, Jaikumar R, Van Wassenhove L. A multiplier adjustment method for control of wind turbines using artificial neural network-Markov chains
the generalised assignment problem. Manag Sci 1986;32:1095–103. prediction with load mitigation. Ain Shams Eng J 2022;13:101577.
[23] Zhang G, Eddy Patuwo B, Y. Hu M. Forecasting with artificial neural networks: [36] Soni A, Koner R, Villuri VGK. Fusion of Dual-Scale Convolution Neural Network
The state of the art. Int J Forecast 1998;14(1):35–62. for Urban Building Footprints. Ain Shams Eng J 2022;13:101622.
[24] Okafor CE, Okafor EJ, Ikebudu KO. Evaluation of machine learning methods in [37] Issa A, Samaneh H, Ghanim M. Predicting pavement condition index using
predicting optimum tensile strength of microwave post-cured composite artificial neural networks approach. Ain Shams Eng J 2022;13:101490.
tailored for weight-sensitive applications. Engineering Science and [38] Hemeida AM, Hassan SA, Mohamed A-A, Alkhalaf S, Mahmoud MM, Senjyu T,
Technology, an International Journal 2022;25:100985. et al. Somaia Awad Hassan, Al-Attar Ali Mohamed, Salem Alkhalaf, Mountasser
[25] Issa A, Samaneh H, Ghanim M. Predicting pavement condition index using Mohamed Mahmoud, Tomonobu Senjyu, Ayman Bahaa El-Din, ‘‘Nature-
artificial neural networks approach. Ain Shams Eng J 2022;13(1):101490. inspired algorithms for feed-forward neural network classifiers: A survey of
[26] Hao Chen, Qian Fu, Qiang Liao, Xun Zhu, Akeel Shah, Applying artificial neural one decade of research”. Ain Shams Eng J 2020;11(3):659–75.
network to predict the viscosity of microalgae slurry in hydrothermal [39] Janmenjoy Nayak, Bighnaraj Naik, H.S. Behera, ‘‘A novel Chemical Reaction
hydrolysis process, Energy and AI, Volume 4, 2021, 100053, ISSN 2666-5468, Optimization based Higher order Neural Network (CRO-HONN) for nonlinear
https://doi.org/10.1016/j.egyai.2021.100053. classification”, Ain Shams Engineering Journal (2015) 6, 1069–1091.
[27] Lawal AI, Kwon S, Hammed OS, Idris MA. Blast-induced ground vibration [40] Nayak SC, Misra BB, Behera HS. Artificial chemical reaction optimization of
prediction in granite quarries: An application of gene expression neural networks for efficient prediction of stock market indices. Ain Shams Eng
programming, ANFIS, and sine cosine algorithm optimized ANN. Int J Min Sci J 2017;8(3):371–90.
Technol 2021;31(2):265–77. [41] Fahmy MMM. Fahmy, ‘‘Online handwritten signature verification system
[28] Bello VE, Olafadehan OA. Comparative investigation of RSM and ANN for based on DWT features extraction and neural network classification”. Ain
multi-response modeling and optimization studies of derived chitosan from Shams Eng J 2010;1(1):59–70.
Archachatina marginata shell. Alex Eng J 2021;60(4):3869–99. [42] Hyndman Rob J, Koehler Anne B. Another look at measures of forecast
[29] Maya R, Hassan B, Hassan A. Develop an artificial neural network (ANN) model accuracy. Int J Forecast 2006;22(4): 679-688.
to predict construction projects performance in Syria, Journal of King Saud [43] Bai G, Tian Z, Zuo MJ. Reliability evaluation of multistate networks: An
University -. Eng Sci 2021. doi: https://doi.org/10.1016/j.jksues.2021.05.002. improved algorithm using state-space decomposition and experimental
ISSN 1018–3639,. comparison. IISE Trans 2018;50(5):407–18.
[30] Al-Fattah SM, Startzman RA, et al. Predicting natural gas production using [44] Schäfer L, Garcia S, Srithammavanh V. Simplification of inclusion–exclusion on
artificial neural network. SPE Hydrocarbon Economics and Evaluation intersections of unions with application to network systems reliability. Reliab
Symposium. Society of Petroleum Engineers; 2001. Eng Syst Saf 2018;173:23–33.
[31] Sagheer A, Kotb M. Time series forecasting of petroleum production using deep [45] Hao ZF, Yeh WC, Wang J, Wang GG, Sun B. A quick inclusion-exclusion
LSTM recurrent networks. Neurocomputing 2019;323:203–13. technique. Inf Sci 2019;486:20–30.
[32] Cinar AC. ‘‘Training Feed-Forward Multi-Layer Perceptron Artificial Neural [46] Bai G, Zuo MJ, Tian Z. Ordering heuristics for reliability evaluation of multistate
Networks with a Tree-Seed Algorithm. Arab J Sci Eng 2020;45(12):10915–38. networks. IEEE Trans Reliab 2015;64(3):1015–23.
[33] Liu Y, Zhao M. An obsolescence forecasting method based on improved radial
basis function neural network. Ain Shams Eng J 2022;13:101775.

10