MATHEMATICAL PROGRAMMING

(MAT707)
TEST 1 PART 2 (45 MARKS)
Oct. 2020 – Feb. 2021
Faculty of
Computer &
Mathematical
Sciences Lecturer: Assoc. Prof. Dr. Adibah Shuib

Question # Total Marks Marks Obtained

Answer ALL questions Question 1: 15 marks
and show all work. Question 2: 15 marks
INSTRUCTIONS
Question 3: 8 marks
Question 4: 7 marks
Total Marks 45 marks

QUESTION 1: LINEAR PROGRAMMING (LP) MODEL FORMULATION (15 marks)

Formulating Transportation Problems

Powerco has three electric power plants that supply the needs of four cities. The amount of supply in
kilowatt-hours (kwh) of electricity from Plant 1, Plant 2 and Plant 3 are 35 million kwh, 50 million kwh
and 40 million kwh, respectively. The peak power demands in these cities, which occur at the same time
(2 p.m.) are 45 million kwh, 20 million kwh, 30 million kwh; and 30 million kwh for City 1, City 2 and
City 3 and City 4, respectively. The costs of sending 1 million kwh of electricity from plant to city depend
on the distance the electricity must travel. The supply, demand and shipping costs are as given in Table
1.

Table 1: Shipping Costs (in $), Supply, and Demand for Powerco
From To Demand Point j Supply (si )
Supply Point i City 1 City 2 City 3 City 4 (in million kwh)
Plant 1 $8 $6 $10 $9 35
Plant 2 $9 $12 $13 $7 50
Plant 3 $14 $9 $16 $5 40
Demand (d j) (in
45 20 30 30
million kwh)

Given the general formulation of a transportation problem as follows:

Let x ij be the number of units shipped from supply point i to demand point j .
m n
Minimize Total Shipping Cost: C=∑ ∑ cij x ij i=1, 2 , … , m; j=1 ,2 , … , n
i=1 j=1
subject to:
n

∑ x ij ≤ si (i=1 , 2 , … , m) (supply constraints – a row for each supply point)

j=1
m

∑ x ij ≥ d j ( j=1 , 2 , … , n) (demand constraints – a row for each demand point)

i=1
x ij ≥ 0 (i=1 , 2 ,… ,m ; j=1 , 2 , … , n)

Formulate a complete Linear Programming (LP) model to minimize the total power supply cost of
meeting each city’s peak power demand. Define all decision variables and write the objective function
and all constraints without the summation notations. (8 marks)

ANSWER:

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 1
Decision Variables:
Let,
x 11 be the number of units shipped from Plant 1 to City 1
x 12 be the number of units shipped from Plant 1 to City 2
x 13 be the number of units shipped from Plant 1 to City 3
x 14 be the number of units shipped from Plant 1 to City 4
x 21 be the number of units shipped from Plant 2 to City 1
x 22 be the number of units shipped from Plant 2 to City 2
x 23 be the number of units shipped from Plant 2 to City 3
x 24 be the number of units shipped from Plant 2 to City 4
x 31 be the number of units shipped from Plant 3 to City 1
x 32 be the number of units shipped from Plant 3 to City 2
x 33 be the number of units shipped from Plant 3 to City 3
x 34 be the number of units shipped from Plant 3 to City 4
(1 mark)
m n
Minimize Total Shipping Cost: C=∑ ∑ cij x ij i=1, 2 , … , m; j=1 ,2 , … , n
i=1 j=1

Minimize Total Shipping Cost

(C)=¿ 8¿ x 11 + ¿ 6 ¿ x12 + ¿ +¿ 9 ¿ x 14 + ¿ 9¿ x 21+ ¿ + ¿13 ¿23 +¿ +¿ 14¿ x 31 +¿ 9¿ x 32 +¿ +¿ 5¿ x 34
(2 marks)
subject to:
¿¿ x 11 +¿ 6¿ x 12 +¿ 10¿ x 13+¿ ¿ x 14 ≤¿ 35¿ (Plant 1 Supply)
¿¿ 9¿ + ¿ 12¿ x 22 +¿ 13¿ x 23+ ¿¿ x 24 ≤¿ 5 (Plant 2 Supply)
¿¿ 14 ¿ +¿ + ¿ 16 ¿ x33 +¿ ¿ x 34 ≤40 (Plant 3 Supply) (4 marks)
¿¿ 8¿ + ¿ 9¿ x 21+¿¿ x 31 ≥ ¿ 45¿ (City 1 Demand)
¿¿ x 12 +¿ 12¿ x 22+ ¿ ≥ ¿ 20¿ (City 2 Demand)
¿¿ x 13 +¿ +¿ ≥¿ 30¿ (City 3 Demand)
¿¿ x 14 + ¿ +¿ 5¿ x 34 ≥¿ 30¿ (City 4 Demand)
x 11, x 12 , x 13 , x 14 , x 21 , x 22 , x 23 , x 24 , x 31 , x 32, x 33 , x 34 (nonnegativity constraint) (1 mark)

b) Four special cases of LP are i) infeasibilty; ii) redundancy; iii) unboundedness and
iv) alternate optimal solutions. Classify each LP model a), b), c) and d) below as any of the
four special cases described above. [Note: No need to show the workout, simply fill in the
blank with your answer]. (3
marks)

Max. Z=x 1+ x 2 Max. Z=x 1+2 x 2 Max. Z=6 x 1+ 4 x 2

s.t. s.t. s.t.
x 1+ x2 ≤150 x 1+ x2 ≤ 20 x1 ≤ 4
x 1+ x2 ≥ 200 2 x1 + x 2 ≤30 2 x 2 ≤ 12
x1 , x2 ≥ 0 x 1 ≤ 25 3 x 1+2 x 2 ≤ 18
x1 , x2 ≥ x1 , x2 ≥ 0
Answer: Answer: Answer:
infeasibility redundancy Alternate optimal solutions

(1 mark) (1 mark) (1 mark)

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 2
c) The following LP model has alternate optimal solutions. Use the graphical isoline
method to determine two optimal solutions. State the general solutions for the
model and determine another optimal solution using the general solution given. (4
marks)
Minimize Z=3 x1 +2 x 2
s.t. x 1−x 2 ≥ 100
1.5 x 1+ x 2 ≥300
4.5 x 1 +3 x 2 ≥ 300
x1 , x2 ≥ 0

Two Optimal solutions using

isoline method:

Optimal solution 1: (1
mark)
x 1=¿ ¿
x 2=¿ 60¿
Z min =¿ 600¿

Optimal solution 2: (1
mark)
x 1=¿ 200¿
x 2=¿ ¿
Z min =¿ ¿

General solution: (1
mark)
( x 1arbitrary, (600 – 3X1)/2)
160 ≤ x 1 ≤200

Optimal solution 3: (1
mark)
x 1=¿ ¿
x 2=¿ 30¿
Z min =¿ 600¿

QUESTION 2: SIMPLEX METHODS (15 marks)

a) Given the following simplex tableau for an LP maximization problem:

-Z x1 x2 S1 S2 A1 RHS
-Z 1 -2-M 0 - 10 - 2M -M 0 30 – 2M
x2 0 1 1 2 0 0 6
A1 0 -1 0 -2 -1 1 2

Based on the given simplex tableau, has the optimal solution been found? If not,
continue the simplex iteration until the optimal solution is found. State the optimal
MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 3
 solution, if exists and the general solution (if alternate optimal solutions occurs).
Otherwise, classify the given LP problem based on any of the four special cases
(unbounded, infeasibility, degeneracy or alternate optimal solutions). (2 marks)
Ans: Optimal solution is found x1=0,x2=6,S1=0,S2=0,A1=2; but Z max=30-2M
(contain artificial variable, thus infeasibility occur.

b) Given the following LP model.

Minimize Z=−2 x1−3 x 2−x 3

subject to:
x 1+ x2 + x 3 ≤ 40
2 x1 + x 2−x 3 ≥10
−x 2+ x3 ≥ 10
x1 , x2 , x3 ≥ 0

i) Convert the given LP problem to standard form by adding slack, surplus and
artificial variables (whichever applicable). (3 marks)

Answer:
Minimize Z=−2 x1−3 x 2−x 3 +0 S1 +0 S2 +0 S 3 + MA 2 + MA 3
subject to:
x 1+ x2 + x 3 +S 1=40 (1 mark)
2 x1 + x 2−x 3 + S2 + A2=10 (1 mark)
−x 2+ x3 + S 3+ A 3=10 (1 mark)
x1 , x2 , x3 ≥ 0

ii) If the model is to be solved using the M technique, rewrite the objective function
in terms of all variables involved. Then, write the initial simplex tableau for the
minimization LP problem in i) based on M-technique and perform ONE iteration
only of the M-technique. (3
marks)
Answer:
To establish the objective function based on the M-technique, let,
A2=10−2 x 1−x 2+ x 3−S 2 (1) (0.5 mark)

A3 =10+ x 2−x 3−S 3 (2) (0.5 mark)

Minimize Z=−2 x1−3 x 2−x 3 +0 S1 +0 S2 +0 S 3 + MA 2 + MA 3

>> Minimize Z=(−2−2 M ) x 1−3 x 2−1 x 3+20 M −Ms 2−Ms 3+20 M

(1 mark)
Thus,
−Z+ (−2−2 M ) x 1−3 x 2−1 x 3+20 M −MS 2−MS 3=−20 M
(1 mark)

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 4
c) Given following LP problem where you need to use two-phase method to solve it.
Minimize Z=x 1−2 x 2
subject to:
x 1+ x2 ≥ 2
x 1−x 2 ≤−1
x2 ≤ 3
x1 , x2 ≥ 0

i) Given the following optimal tableau of Phase I, express Z , the objective function
of Phase 2 in terms of x 1and x 2. (3 marks)

Optimal tableau of Phase I:

−W x1 x2 S1 S2 S3 A1 A2 RHS
−W 1 0 0 0 0 0 1 1 0
−1 1 1 −1 1
x1 0 1 0 0
2 2 2 2 2
−1 −1 1 1 3
x2 0 0 1 0
2 2 2 2 2
1 1 −1 −1 3
S3 0 0 0 1
2 2 2 2 2

Answer:
Since W* = 0, solution is optimal (End of Phase I), drop all artificial variables A 1
and A2. Express Z in terms of x 1 and x 2,
x 1−¿ (1) (0.5 mark)

3
x 2−1/2 S 1−1/2 S 2= (2) (0.5 mark)
2

So:
x 1=1/2 S 1−¿ (3) (0.5 mark)

x 2=1/2 S 1+¿ (4) (0.5 mark)

and, based on the given LP model, Z=x 1−2 x 2

by substituting,
Z=¿ (1 mark)

ii) Determine the optimal solution of the given LP problem by completing the following
Phase II tableau.
(4 marks)
Phase II
−Z x1 x2 S1 S2 S3 RHS
−Z 1 0 0 -1/2 -3/2 0 5/2 (1 mark)
−1 1 1
x1 0 1 0 0
2 2 2

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 5
 −1 −1 3
x2 0 0 1 0
2 2 2
1 1 3
S3 0 0 0 1
2 2 2
−Z 1 3 0 −¿2 0 0 4
S2 0 2 0 −¿1 1 0 1
x2 0 1 1 −¿1 0 0 2
S3 0 −¿1 0 1 0 1 2
−Z 1 0 0 -1/2 -3/2 0 5/2
S2 0 1 0 0 1 1 2
x2 0 0 1 0 0 1 3
S1 0 0 0 1/2 1/2 1 5/2

Optimal solution: Z=¿ ¿ ; x 1=¿ x 2=¿ ¿ ; S1 =¿ ¿; S2 =¿ ¿; S3 =¿ ¿ (1 mark)

QUESTION 3: SOLVING LP USING EXCEL SOLVER AND REVISED SIMPLEX

METHOD (8 marks)

a) Syntron, a television manufacturer, has decided to produce and sell two different
types of TV sets, TV Product 1 and TV Product 2. They assume that TV Product 1
will give a profit of RM300 per unit and TV Product 2 a profit of RM500 per unit.
Syntron has one production plant with four departments: moulding, soldering,
assembly and inspection. Each TV set is processed in sequence through these four
departments. Each department has a limited capacity given by a maximum number
of working hours per year. We assume Syntron can sell all the TV sets they are
able to produce (the market is not a restriction). The Linear Programming (LP)
formulation of the problem is as follows:
x 1=¿number of TV Product 1 produced per year
x 2=¿number of TV Product 2 produced per year

Maximize Profit, Z=300 x1 +500 x 2 (3.0)

Subject to:
x 1+ 5 x 2 ≤ 4000 (Moulding hours) (3.1)
x 1+ x2 ≤1200 (Soldering hours) (3.2)
2 x1 + x 2 ≤2000 (Assembly hours) (3.3)
2 x1 +5 x 2 ≤ 5000 (Inspection hours) (3.4)
x1 ≥ 0 ; x2 ≥ 0

Excel Solver Answer Report and Sensitivity Report for the LP problem is given
below:

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 6
 Answer Report:

Sensitivity Report:

i) Based on the solution, are there unused working hours? If yes, state in which
department and how many hours are unused? (1 mark)
Answer:
Assembly and Inspection

ii) What is the range of feasibility of moulding constraint? Will the shadow price
change if the Right-Hand Side (RHS) of moulding constraint is changed to
3000? Why? (2
marks)
Answer:
(-∞,4000] the shadow price will not change but the objective function will change.

b) Given an LP model in standard form and converted to equality constraints as follows:

LP Standard Form with Inequaliti Constraints

LP Model in Standard Form
Converted to Equality Constraints

Maximize Z=4 x 1 +3 x 2+ x 3 +7 x 4 +6 x 5 Maximize Z=4 x 1 +3 x 2+ x 3 +7 x 4 +6 x 5 + x 6 + x 7+ x 8

subject to: subject to:
x 1+ 2 x 2 +3 x3 + x 4 −3 x 5 ≤ 9 x 1+ 2 x 2 +3 x3 + x 4 −3 x 5+ x 6 =9
2 x1 −x2 +2 x 3+ 2 x 4 + x 5 ≤10 2 x1 −x2 +2 x 3+ 2 x 4 + x 5 + x 7=10
−3 x 1+ 2 x 2 + x 3−x 4 +2 x5 ≤ 11 −3 x 1+ 2 x 2 + x 3−x 4 +2 x5 + x 8=11
x1 , x2 , x3 , x4 , x5 ≥ 0 x 1 , x 2 , x 3 , x 4 , x 5 , x 6 , x 7 , x5 ≥ 0

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 7
Perform the Step 4, 5 and 6 of the first iteration ONLY of the Revised Simples
Method (RSM) to determine the optimal solution of the model.
(5 marks)
Answer:

First Iteration:

[ ]
1 0 0
−1
Step 1: Initial Step  Set B = 0 1 0 =I
0 0 1

[]
and define C B
x1
x2
x

[ ] []
1 2 3 1 −3 1 0 0 3 9
x
2 −1 2 2 1 0 1 0 4 = 10
−3 2 1 −1 2 0 0 1 x 5 11
x6
x7
x8

[][ ]
x5 9
T
 X B=[ 6 7 8 ]
x x x = x7 = 10
x8 11

[ ][]
x1 0
x2 0
T
X N =[ x1 , x 2 , x 3 , x 4 , x5 ] = x 3 = 0 ;
x4 0
x5 0
C B= [ 0 ,0 , 0 ] i.e., coefficients of x 7 , x 8 , x9 ; Thus, C B B−1=[ 0 , 0 , 0 ]

Step 2: Determine entering variable.

Z j−c j=Max ( Z j−c j )=Max j ( C B B−1 P j −c j )=( 1 , C B B

−1
) ( )
−c N
Pj
; j non basic index.
If Z j−c j ≥ 0 ∀ j , solution is optima; and go to Step 6. Otherwise, choose x s as the
entering variable and go to Step 3.

[ ]
−4−3−1−7−6
Z j−c j=( 1 , C B B
−1
) ( )
CN
Pj
= [ 1 0 0 0]
1 2 3 1−3
2 −1 2 21
−3 2 1−1 2

= [ −4−3−1−7−6 ]

j = 1, 2, 3, 4, 5  x 4 enters the basis ( x s=x 4 ) since the most negative (-7) for –Z row is
at j = 4.

Step 3: Which variable to be leaving from the basis?

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 8
 () [ ][ ] [ ] [ ][ ] [ ]
a64 1 0 0 1 1 1 0 0 9 9
( a ks ) = a74 = B−1 P 4= 0 1 0 2 = 2 ; B−1 P 0= 0 1 0 10 = 10 ,
a84 0 0 1 −1 −1 0 0 1 11 11
K = 6, 7, 8 (basic variable)

{ } {91 , 102 , −111 }=Min {9 , 5 ,−11 }=5

−1
B P0
such that ( B P6 ) k > 0
−1
θ=Mink −1
i.e., Mink
B P6
 (( x r=x 7 will be leaving variable,r =7)

Step 4: Determine B−1

B
−1
= E. B−1 where E = (e 6, K , e 8) with position K as r = 7

() [ ][ ] [ ]
a64 1 0 0 1 1
Recall, ( ks ) a74 = B P 4= 0 1 0 2 = 2 k = 6, 7, 8 (basic var), s=4 .
a = −1

a84 0 0 1 −1 −1

() ()
−a ks −a 64
ars a 74

( )
1 1 −1 /2
K= ; k = 6, 7, 8; r = 7; s = 4 K= = 1/2 (0.5 mark)
ars a 74
1/2
−a ks −a 84
ars a 74

Compute B−1

[ ][ ] [ ]
1 −1 /2 0 1 0 0 1 −1/2 0
−1 −1
B =¿ E. B = 0 1 /2 0 0 1 0 = 0 1/2 0 ; (0.5 mark)
0 1 /2 1 0 0 1 0 1/2 1

[]
x6
T
X B=[ 6 4 8 ]
x , x , x = x 4 ; thus, C B= [ 0 7 0 ] objective function’s coefficients of x 6 , x 4 , x 8
x8

coefficient of x 4 takes the place of the coefficient of x 7; i.e (xs = x 4 ; xr = x 7), s=enter,
r=leaving.

Step 5: Compute C B B−1

[ ]
1 −1/2 0
−1
 Thus, C B B =[ 0 7 0 ] 0 1/2 0 =[ 0 7/2 0 ] (1 mark)
0 1/2 1

Step 6: Important Information

[]
9
−1
C
Current obj function: X0 = B B P 0 = [ 07 /2 0 ] 10 =35 ; (0.5 mark)
11

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 9
 [ ][ ] [ ]
1 −1 /2 0 9 4
T
Current solution: X B=[ x 6 x 4 x 8 ]
−1
= B P 0=¿ 0 1 /2 0 10 = 5
0 1 /2 1 11 16
x 6=4; x 4 =5, x 8=16 (1.5 mark)

QUESTION 4: FUNDAMENTAL THEOREM OF LP, WEAK DUALITY THEOREM,

SECOND DUALITY THEOREM AND COMPLEMENTARY SLACKNESS THEOREM
OF LINEAR PROGRAMMING (7 marks)

a) Given a Linear Programming (LP) problem in 8 variables and 13 constraints,

determine the upper bound for the number of basic feasible solutions. (Hint: Think
of the dual) (1
mark)
Answer:

b) Prove the following Weak Duality Theorem:

Weak Duality Theorem:
¿ ¿
Let x be a feasible solution to the primal and let y be a feasible solution to the
T
dual, then c T x ¿ ≤ ( y ¿ ) b (2
marks)

Answer: Proof:
Primal (P) Dual (D)
T T
max c x min y b
s.t. Ax ≤ b s.t. ATy ≥ c
x≥0 (1) y≥0 (2)
¿
Let y be a feasible solution to the Dual.
¿
Based on (2): AT y ≥ c
 c ≤ ________ (0.5 mark)
T
 c ≤ ________ (0.5 mark)
 cT ≤ (y*)T A
Multiply both sides by x*
 cTx* ≤ ______ (0.5 mark)

But Ax* ≤ b from (1.0)

Thus, cTx* ≤ ___________ ≤ (y*)Tb (0.5 mark)
or simply: cTx ≤ yTb

c. Complete the following statement. (1 mark)

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 10
 The Strong Duality Theorem:
If either P or D has a finite optimal value of objective functions, then so does
the other, these optimal values coincide, and optimal solutions to both P and D
exist.
Based on the Strong Duality Theorem, one of the possibilities is
If either P or D has an unbounded value of objective function, then the other problem
___________________________.

d) Write the Primal (P) formulation for the given LP Dual (D) as the following.

SOLUTION:
Dual (D) Problem Primal (P) Problem

Maximize W =3 y 1+ 8 y 2+2 y 3 Minimize Z=5 x1 +6 x 2−x 3

subject to: subject to:
y 1 + y 2=5
y 1 +2 y 2+ 4 y 3 ≤6
3 y 2 +5 y 3 ≤−1
y 2 ≤0 , y 3 ≥0

¿ 16 −1
Given the optimal solution of the dual (D), y =[ y ¿ ¿ 1 , y 2 , y 3 ]=[ 0]¿ and for the
3 3
T
5
primal (P), x ¿=[ x ¿ ¿1 x 2 x 3 ]T =[3 0 ] ¿, what are the objective function values for
3
both D and P. Show that the Strong Duality Theorem is satisfied.
(3 marks) Answer:

Honour Pledge for MAT707 Test 1 Part 2

The Honor Pledge is a community building ritual, designed to encourage faculty and
students to reflect upon the University's core institutional value of academic integrity.
For students, it is a statement that students should be asked to write by hand and sign
on examinations papers, or other academic assignments. For this online Test 1 of
MAT707, you will be asked to type your full name and student ID as an agreement that
you fully understand what it means to conduct yourself with academic integrity. This is
also a show of commitment to honesty.

* Honour Pledge for Test 1 Part 2:

“I affirm that I have not given or received any unauthorized help on this
Test 1 Part 1 (Online), and that all works done and answers given are my
own. I will be responsible for any violation of this pledge.”

Name: Hanisah bt Azhar

Student ID: 2020692134

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 11
 Date: 11/12/2020

* Note: To sign, write your Full Name (Your Name followed by Your Father's Name),
your e-mail address and date below.

MAT707 Mathematical Programming, Assoc. Prof. Dr. Adibah Shuib, FSKM, UiTM 12