The simplex method requires that all constraints be expréssed as equations. Mathematically,
it is easier to solve system of linear equations than systems of inequalities. Therefore, all the
inequalities shall be converted into equations or in the standard form of linear programming.
problem. Simplex Method is an iterative technique that begins with a feasible solution that is not
optimal, but serves as a starting point. With the use of algebraic manipulation, the solution is
improved until no further improvement is possible. It is more convenient to solve linear
programming models in simplex method with more two unknown variables because
geometrically it is difficult to graph. The idea of the simplex method is to start at the corner
where all the unknowns are 0 and then walk around the region from one corner to another the
value of the objective function always increases (for maximization) and always decreases (for
minimization) until the best corner is determined.
, ‘Thesolutionis
[aes "optimal, >] End
Set up the initial
tableau
2
Arethere negative
entries in the last row?
Determine the pivot
column,
‘Arethere positive
entriesin test ratio?
No
(No finitesolution
CDP exists. (>| ad
Determine the pivotal
+ row.
Figure 3.1: Simplex Method Flowchart
Computes anew
tableau,
Figure 3.1 shows the simplex method flowchart for LP maximization model and is not
applicable to all LP model. ‘There is a slight difference in establishing the solution of an LP
minimization model. This will further be discussed in the later section of the text.
Before we go into the solution using simplex method we have to determine first the general
and standard linear programming models.
A. General Linear Programming Model
A linear programming problem in n unknowns 2:, x2, %3,..., Xx is gne which we are to
determine the maximum or the minimum value of the objective function,
1x1 + axe + asx +...+ aan, Where Mi, M2, 43,..., ax are constants, subject to a linear constraints of
the form :
bux + boxe + baxs +... Dux SC OF
bux + boxe + baxa +..+ bath 2.6 OFbua + boxe + bara +... Datn =e
where bi, bz, bs,...,bn and c are numbers.
B. Standard Linear Programming Maximization Model
A standard linear programming maximization model are required to maximize an objectiy
function of the form
Maximize: P = arxi + aox2 + a3x3 +...+ ann
Subject to: bixi + baxa + bars +...4 bat $c
x1 20,2220, x320,...,2n20
where c > 0 and the inequality in the structural constraints is strictly “<”.
structural constraints
non-negativity constraints
C. Standard Linear Programming Minimization Model
A standard linear programming minimization model are required to minimize an objectiy
function of the form
Minimize: C= aixi + aaxa + ass +...+ dan
Subject to: bixi + boxe + baxs+...+ bun 2 € structural constraints
m1 20,2220, x320,...,xn20 non-negativity constraints
where c > 0 and the inequality in the structural constraints is strictly “>”.
# Enrichment Exercise 3.1
Determine whether the following are standard LP model or non-standard LP model.
1. Minimize: C= 13x: + 18x2+ 22x3 3. Minimize: C = 12x: + 2x2+ 18x3
Subjectto: x1+x2 +432 370 Subject to: w+ 5x3 265
6x1 + 8x2 > 480 321 + 5x2 = 20
2u1+ xt x92 230 22+ 7x9 < 80
x120,x220,x220 x120,2220,x320
2. Maximize: P = 15x1 + 33x2+ 18x3 4. Maximize: P = 9x1 + 6x2+ 3x3
Subject to: w+ x2 + x3 $355 Subject to: X14 x2+ x32 30
21+ 4x2 + 5x3< 420 7x1 + x24 4x3 > 30
2xi+ 2xa+ x3 < 500 x2t 2x3 > 30 i
x12 0, 2220, x32>0 2120, 2220, x320 |
We use the same example presented in Chapter 2 (see Page 33). The following steps wil
Buide us in solving standard linear programming maximization problem. The maximality test of
simplex method in a maximization problem triggers an optimal solution if and only if the la
row of a simplex tableau, corresponding to the objective function, contains no negative entries.implex Method
del using Si
1, Set up problem in an LP model. ;
2. Introduce the necessary slack variables,
3. Establish the initial tableau. “
4, Examine the simplex tableau for optimal solutions, If the basic feasible solution is maximal,
the problem is solved. Otherwise, proceed to step 5.
5, Compute a new simplex tableau: select the pivot column, find pivot row, and pivot about |
the pivot entry. |
| 6. Proceed to step 4. J
|
Example: A local boutique produced two designs of gowns A and B and has the following
materials available: 18 square meters of cotton, 20 Square meters of silk, and 5 square meters of
wool.
ign A requires the following: 3 square meters of cotton, 2 square meters of silk and
square meter of wool. Design B requires the following: 2 square meters of cotton, 4 square meters
of silk. If Design A sells for P1,200 and Design B for P1,600, how many of each garment should
the boutique produce to obtain the maximum amount of money?
Solution:
In order to solve a standard linear programming maximization model using simplex method
it is necessary to follow the steps.
tsp I: Represent the unknown in the problem.
Let x: be the number of Design A gowns, and
xzbe the number of Design B gowns. -
Step 2 Tabulate the data about the facts (if necessary).
Materials | Design A(x) | Design B(x) | Available
3
Cotton 2 18
Silk 2 4 20
Wool - ' 1 0 5
Profit P1,200 P1,600
Step 5: Formulate ‘the objective function and constraints. by restating the information in
mathematical form.
The objective function is:
Maximize: P = 1,200x1 + 1,600%2
The constraints are:
3x1+2m2<18 => Cotton,
2mt+42<520 => Silk Structural Constraints
ms5 => Wool
m20,x220 Non-negativity ConstraintsStep # Convert to a system of linear equations. Take note that 3x1 + 2x2 < 18 are less convenien
than equations, We have to adda slack variable denoted by Sn, which satisfies the
condition of $: 2-0. This is known as standard form, it is a way of expressing th
constraints of a Linear Programming problem as equalities with all variables on the lej
side of the equation and a constant on the right side. Thus, we can say that slac)
variables are variables added to constraints to convert them into equations.
Let Su represents thé slack variables.
3x1 + 2x2,4,51 = 18
2xi + 4x2 + $= 20
mtS=5
~1,200e1 = 1,6004+ P= 0
or it may be written as
3x1 + 2x2 + $1 +02 + 0S3 + OP =18
2x1 + din + 0S: + S24 0Ss + OP = 20
21 +051 +052 + $3 + 0P =5
1,200xi = 1,600x2 + 0S: + 0S: + 0S3+ P= 0
1 Constraint
2-4 Constraint
3" Constraint
Objective Function
1* Constraint
2°4 Constraint
3 Constraint
Objective Function
Step 5: Set up the initial tableau.
Tableau
BV | mm Si. So Ss P| RHS
St 3 2 1 0 0 0 18 | 1* Constraint
S 2 4 0 1 0 0 20. | 2-4 Constraint
Sa 1 0 ‘0 0 1 0 5__| 3 Constraint
P__|-1200 16000 0 0 i 0 | Objective Functior
Beep & Select the pivot column. (It is the column that contains the most negative entry in the
. bottom row.) In this example itis the column of x: which contains —1,600.
Tableau
BY [om x Si Se Ss P| RHS
St 3 2 1 0 0 0 18
S 2 4 0 1 0 0 20
S35 0 0 0 1 0 5
P 1,200. 600 0 0 0 1 0
aay
Iteration is a simplex method which’ consists of
performed in moving one basic feasible solution to
keep track of the calculations made when the simp!
(RHS) is the column in a simplex tableau indicating the
Basic Variables (BV) are the variable included in a basic
the sequence of steps (row operations)
another. Simplex Tableau is a table use to
x method is employed. Right-Hand-Side
quantities of the variables is in a solution.
solution.Pivot column is the column in any solution to a maximization problem which has the lowest
negative value in the last row. Intersectional Elements are elements common to both the pivot
column and the rows representing variables in the solution. Pivot Row is the row in the simplex
tableau corresponding to the basic variable that will leave the solution. It is determine by the test
ratio and it is being computed by dividing the right-hand-side (RHS) by the intersectional
elements (IE). Pivot is the element of the simplex tableau that is in'both the pivot row and the
pivot column. The test ratio. result always be a positive number (meaning negative number and
infinity cannot be chosen as pivot row). After applying the test ratio choose the smallest positive.
In cases, that two or more candidates choose any one. This process of going from one simplex
tableau to the next is called pivoting. : y
tsp 7: Identify the pivot row using the test ratio by dividing the RHS values by the non-zero
and non-negative entries in the pivot column (or intersectional element).
The second row of the constraints coefficient has the smallest positive quotient, which is
5; second row will be the pivot row. ;
Tableau
By | on x2 Si. Se ‘Ss. P__|_-RHS | Test ratio
St Bop ne 1 0 0 0 18 | 18+2=9
Sz 20+4=
S 1 0 0 0 1 0 5 | 5+0=not permissible
P_ _[-1,200 .-1,600 0° 0 0 1 0
Tableau
BV [xm x2 Si Se 3 P__| RHS
Sn 3. 2 a 0 0 0 18
Se 2 a 0 1 0 0 20 ;
Ss 1 0 0 0 1 0 5
P| -1200 “-1,600 0 0 0 1 0
Other parts of the simplex table are:
= a
BV 2 31 Se Sa P._| _RHS
St 3 1 0 0 0 18 .
2 0 1 0 0 20
$ [4 0 0 eso) 5
P [=1200 “=1,600\ 0 0 0 a 0
Then initiate the following steps.Tableau
By [om 2 St & Ss P__|_RHS
Si 3 2 1 ORs =0) 0 18
Sf 2a 0 1 0 0 20
Ss 1 0 0 0 1 0 5
P_[-1200- 1,600 "0 0 0 1 0
Before we start with the computation for the solution, let us be aware of the following
5"
& Compute the values of the replacing row by dividing all the entries by the pivot 4,
8 ing,
Replacing Row = Pivot Row + Pivot
R2= Rol + Pa
R2=(2,4,0,1,0,0, 20) +4
* BP=(%, 1,0, %,0,0,5)
—_——_
Tee]
th
% Compute the new values for the remaining rows using the formula.
iB iB
«Ri
Rt
eR!
<—Re
—_
Tira]
. Remaining Row = Previous Row - (Intersectional Element x Replacing Row)
Re=Ri-
R?= (3,2, 1,0, 0, 0, 18)—{2(*%, 1, 0, %,.0, 0, 5)
R= (3,2, 1,0, 0, 0, 18) - (1, 2, 0, %, 0, 0, 10)
2R2
“Ri =(2,0, 1, -%, 0, 0, 8)
2 = Rs
R#= (1, 0,0, 0,1, 0,5) - (0), 1, 0, %, 0, 0, 5)
OR?
R#= (1,0, 0, 0, 1, 0,5) - (0, 0, 0, 0, 0, 0,0)
“R# = (1,0, 0, 0, 1, 0, 5)
Re= Ra + 1600R2
Re = (-1200, -1600, 0, 0, 0, 1,0) + (1600)(%, 1, 0, %, 0, 0, 5)
Ré = (-1200, -1600, 0, 0,,0, 1, 0) + (800, 1600, 0, 400, 0, 0, 8000)
Re = (—400, 0, 0, 400, 0, 1, 8000)
Enter the results in Tableau 2.
Tabloau2 :
AV. [ae ge gay P| RES
af eee ae 11a 0 0 8
is % 1 0 % 0 0 5
Ss 1 0 0 0 aL 0 5
Lp [400.0 o - 40° 0 1__[ 8,000
RB
«R22
< RP
Re.[a Note: The pivot entry transforms to 1 and all other entries in the pivot column transforms — |
|
[esgs ek WO eroa” oe ee
Return to Step 6, since the last row still contains a negative entry.
Step 7: Select the pivot column, identify the pivot row and the pivot of Tableau 2.
Tabloau2
BV x x2 St Sa Ss P RHS | Test ratio
St T 0 i h 0 0 8 | 8+2=
na | % 1 6 op 0 5 |5+%=10
Ss 1 0 0 0 1 0 5 [5+1=5
P| Mo 0. 400 ~O 1 | 8,000
Step 8-9 Compute for replacing row and the remaining rows of Tableau 3.
Tableau 2
BV x x2 Si So Sa P RHS
St 2 0 t an 0 0 & «Re
n | % 1 0 We0 0 5 ler
Ss 1 0 0 0 £ i) e << R?
P| 4000 0 4000 1 [ 8000 | «Re
Replacing Row = Pivot Row + Pivot
Re=Re+ Pe
Rx = (2,0, 1,-%, 0, 0,8) +2
Re= (1,0, %—%, 0, 0, 4) “
Rehpining Row = Previous Row = {intersectional Element x Replacing Row)
= R2-2R8
pe (%, 1, 0, %, 0, 0, 5) — PA)(L, 0, ¥% -¥, 0, 0, 4)
R= (4,1, 0, % 0, 0,5) ~ (4, 0, YM, 0, 0,2)
RP=(0, 1,-%, % 0,0,3)
RS= Re-ORP
R= (1, 0,0,0, 1, 0, 5) ~ (1)(4, 0, % 0,0, 4)
Rs =(1, 0, 0, 0,1, 0,5) — (1, 0, % Ys, 0, 0,4)
RS = (0, 0,-¥%, %, 1, 0, 1)
Ré = Re + 1600?
Ré = (-400, 0, 0, 400, 0, 1, 8000) + (400)(1, 0, 14; -¥% 0, 0, 4)
Re = (400, 0, 0, 400, 0, 1, 8000) + (400, 0, 200, -100, 0, 0, 1600)
R&= (0, 0, 200, 300, 0, 1, 9600)
Enter the results in Tableau 3.Tableau 3
BY. [an x Si Sa So P__|_RHS
x 1 0 % —" 0 0 4 |