Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 11

IE 409 PROJECT MANAGEMENT AND SCHEDULING

INDIVIDUAL PROJECT

EVALUATION OF ALTERNATIVE PROJECTS IN

TURKISH DEFENSE INDUSTRY

1
Table of Contents

1. Part A : Mathematical Modelling ………………………………………………..………….1

2. Part B : Additional Constraints

……………………………………………………………..3

3. Part C : Trial Study For The Final Bid ………………………………………………………

References ……………………………………………………………………………………..6

Appendix ………………………………………………………………………………………7

Table of Figures

Figure 1: Solution matrix for project selection and initiation

…………………………………..3

Figure 2 : Solution statistics ……………………………………………………………………

Figure 3: Solution matrix with additional constraints

………………………………………….4

Figure 4: Solution statistics with additional constraints

………………………………………..4

Figure 5: Network diagram

…………………………………………………………………….5

2
Figure 6: Excel calculations ...

………………………………………………………………….5

3
Part A: Mathematical Modelling

Mathematical modelling of the problem is given below.

Sets and Indices

PROJECTS = {1,2,3,4,5} -> i

YEARS = {1,2,3,4,5,6} -> t

PDUR = {1,2,3,4} -> r(This is defined in order to consider year based costs)

Parameters

bi : Expected benefit scores of project i.

ft : Available budget of year t.

di : Duration of project i.

cir : Expected cost of project i in year r.

pl : Set of precursors for project l.

Decision Variables

xit = 1, if the project i starts on year t

0, otherwise

Objective Function

5 6
MaxZ = ∑ . ∑ bi . xit
i=1 t =1

1
Constraints

∑ xit ≤ 1, i=1,2,3,4,5
t =1

5 t

∑ . ∑ c ( i, t +1−r ) . xir ≤ ft ,t=1,2,3,4,5,6


i=1 r=1

∑ t . xit + Di ≤ 7 , i=1,2,3,4,5
t =1

x11 = 1

6 6

∑ xit ≥ ∑ xlt ,i ∊ Pl
t =1 t=1

∑ t∗xlt+ 7∗¿ ¿
t =1

Solution of The Mathematical Model

The model formed has been solved in IBM CPLEX optimization software. As a result

of computing of the CPLEX optimum value has been found to be MaxZ = 1090 with x11 = 1,

x31 = 1, x44 = 1, x53 = 1 which means Project 1 and 3 should be initiated in the beginning of

year 1 and Project 5 should be initiated in year 3 and Project 4 in year 4. In the calculations of

the software Project 2 has not been selected to employ to gain the optimum value of Z.

Figures about the solution are given on the next page and CPLEX codes and DAT codes are

given in the Appendix. Rule of precursion is also not violated as Project 4 starts after Project

3 ends.

2
Figure 1: Solution matrix for project selection and initiation

Figure 2 : Solution statistics

Part B: Additional Constraints

In order to determine whether there are one or more alternative optima below

constraints have been added to model.

Sk = set of selected projects at k (original optimal solution).

Bk = number of projects selected at k.

3
5 6

∑ . ∑ bi . xit =Z∗¿
i=1 t =1

5 6

∑ . ∑ xit < Bk , xit ∊ Sk


i=1 t =1

When the model is solved again with additional constraints a new optimum solution is

discovered with MaxZ = 1090 again with x11 = 1, x31 = 1, x44 = 1, x52 = 1 which means Project

1 and 3 should be initiated in the beginning of year 1 and Project 5 should be initiated in year

2 and Project 4 in year 4. In the calculations of the software Project 2 has not been selected to

employ to gain the optimum value of Z. Figures about the solution are given on the below and

additional constraints are also given in the CPLEX codes and DAT codes are given in the

Appendix.

Figure 3: Solution matrix with additional constraints

Figure 4: Solution statistics with additional constraints

4
Part C: Trial Study For The Final Bid

In the final section of the project bid values are calculated. Results and network

diagram are given below.

Figure 5: Network diagram

Figure 6: Excel calculations

Excel calculations reveal that earlier payments incur less total costs although

contractors tend to start activities as late as possible. Bid 1 has total estimated cost of

5
35.649,72 USD while Bid 2, which should be selected, has 37.956,01 which signals to nearly

2000 USD difference.

References

Souder, W. E. (1973). Analytical effectiveness of mathematical models for R&D project

selection. Management science, 19(8), 907-923.

Mohagheghi, V., Mousavi, S. M., Vahdani, B., & Siadat, A. (2017). A mathematical modeling approach

for high and new technology-project portfolio selection under uncertain environments. Journal of

Intelligent & Fuzzy Systems, 32(6), 4069-4079.

Naderi, B. (2013). The project portfolio selection and scheduling problem: mathematical model and

algorithms.

Turner, J. R., & Simister, S. J. (2001). Project contract management and a theory of

organization. International journal of project management, 19(8), 457-464.

Turner, R. (2017). Farsighted project contract management. In Contracting for Project

Management (pp. 33-57). Routledge.

Reza Hoseini, A., Ghannadpour, S. F., & Hemmati, M. (2020). A comprehensive mathematical model

for resource-constrained multi-objective project portfolio selection and scheduling considering

sustainability and projects splitting. Journal of Cleaner Production, 269, 122073.

6
Appendix

MODEL CODES
//Sets and Indices

int i = ...;
range PROJECTS = 1..i;

int t = ...;
range YEARS = 1..t;

int r = ...;
range PDUR = 1..r;

int z = ...;
int bk = ...;

//Parameters

float b[PROJECTS] = ...;


float f[YEARS] = ...;
float d[PROJECTS] = ...;
float c[PROJECTS][PDUR] = ...;
int p4 = ...;

//Decision Variables

dvar int+ x[PROJECTS][YEARS] in 0..1;

//Objective Function

maximize sum(i in PROJECTS, t in YEARS) b[i]*x[i][t];

//Constraints

subject to {

sum(i in PROJECTS, t in YEARS) b[i]*x[i][t] == z;

(x[1][1]+x[3][1]+x[3][2]+x[3][3]+x[3][4]+x[4][1]+x[4][2]+x[4][3]+x[4][4]+x[5]
[1]+x[5][2]+x[5][3]) <= bk;

97*x[1][1]+110*x[2][1]+155*x[3][1]+37*x[4][1]+108*x[5][1] <= 280;


86*x[1][1]+110*x[2][2]+192*x[2][1]+155*x[3][2]+129*x[3][1]+37*x[4][2]+30*x[4]
[1]+81*x[5][2]+108*x[5][1] <= 340;
110*x[2][3]+192*x[2][2]+143*x[2][1]+155*x[3][3]+129*x[3][2]+119*x[3][1]+37*x[4]
[3]+30*x[4][2]+36*x[4][1]+108*x[5][3]+81*x[5][2]+89*x[5][1] <= 300;

7
192*x[2][3]+143*x[2][2]+156*x[2][1]+155*x[3][4]+129*x[3][3]+119*x[3][2]+37*x[4]
[4]+30*x[4][3]+36*x[4][2]+81*x[5][3]+89*x[5][2]+65*x[5][1] <= 300;
143*x[2][3]+156*x[2][2]+129*x[3][4]+119*x[3][3]+30*x[4][4]+36*x[4][3]+89*x[5]
[3]+65*x[5][2] <= 280;
156*x[2][3]+129*x[3][5]+119*x[3][4]+36*x[4][4]+65*x[5][3] <= 275;

forall(i in PROJECTS)
sum(t in YEARS) x[i][t] <= 1;

forall(i in PROJECTS)
sum(t in YEARS) t*x[i][t]+d[i] <= 7;

x[1][1] == 1;

sum(t in YEARS) (x[3][t]) >= sum(t in YEARS) x[4][t];

sum(t in YEARS) (t*x[4][t]) + 7*(1-sum(t in YEARS) (x[4][t])) - sum(t in YEARS)


(t*x[3][t]) >= d[3]*sum(t in YEARS)(x[3][t]);
}

DATA CODES
i = 5;
t = 6;
r = 6;
z = 1090;
bk = 4;
b = [220 210 250 240 380];
f = [280 340 300 300 280 275];
d = [2 4 3 3 4];
p4 = 3;
c = [[97 86 0 0]
[110 192 143 156]
[155 129 119 0]
[37 30 36 0]
[108 81 89 65]];

You might also like