Linear Programming




                Terminology
What is a Mathematical Model ?
F=ma
              ‘Mathematical Expressions’

o   Here m and a are called as ‘Decision Variables’


o   F can be called as ‘Objective Functions’


o   Now, F can be controlled or restricted by limiting m or
    a … say m < 50 kg …here, m can be called as a
    ‘Constraint’


o   Similarly if a > o …always, then this condition is called
    as ‘Non-Negativity Condition’

                     http://www.rajeshtimane.com              2
Illustration:
Maximize: Z = 3x1 + 5x2                         Objective
                                                Function

Subject to restrictions:
                   x1                  <4
                                                Functional
                   2x2                 < 12     Constraints
                   3x1 + 2x2           < 18

Non negativity condition
                 x1                     >0      Non-negativity
                 x2                     >0      constraints

                  http://www.rajeshtimane.com               3
What is Linear Programming (LP)?

 The most common application of LP is allocating
  limited resources among competing activities in a
  best possible way i.e. the optimal way.

 The adjective linear means that all the mathematical
  functions in this model are required to be linear
  functions.

 The word programming does not refer to computer
  programming; rather, essentially a synonym for
  planning.



                  http://www.rajeshtimane.com         4
Graphical Solution
Ex) Maximize: Z = 3x1 + 5x2

Subject to restrictions:
           x1 < 4
           2x2 < 12 i.e. x2 < 6
           3x1 + 2x2 < 18

Non negativity condition
          x1, x2 > 0

Solution: finding coordinates for the constraints (assuming perfect equality), by putting
one decision variable equal to zero at a time.


Restrictions (Constraints)        Co-ordinates

x1 < 4                            (4 , 0)

x2 < 6                            (0 , 6)

3x1 + 2x2 < 18                    (0 , 9) & (6 , 0)


                                  http://www.rajeshtimane.com                               5
Restrictions (Constraints)       Co-ordinates                      Non-negativity Constraint

x1 < 4                           (4 , 0)                                             x1, > 0
x2 < 6                           (0 , 6)                                              x2 > 0
3x1 + 2x2 < 18                   (0 , 9) & (6 , 0)




     X2


     10


         8

             A       B
         6


         4
                             C       Feasible Region (Shaded / Points A, B, C, D and E)
         2


         0                   D
             E   2       4       6         8    10         X1

                                     http://www.rajeshtimane.com                               6
Feasible Solutions

 Try co-ordinates of all the corner points of
  the feasible region.

 The point which will lead to most
  satisfactory objective function will give
  Optimal Solution.

 Note: for co-ordinates at intersection; solve
  the equations (constraints) of the two lines
  simultaneously.

                 http://www.rajeshtimane.com     7
Optimal Solution
Corner   Limiting Constraint             Co-ordinate         Max. Z= 3x1 + 5x2


   A     x2 = 6                            (0 , 6)                  30

   B     x2 = 6 & 3x1 + 2x2 = 18           (2 , 6)                  36

   C     x1 = 4 & 3x1 + 2x2 = 18           (4 , 3)                  27

   D     x1 = 4                            (4 , 0)                  12

   E     Origin                            (0 , 0)                  0




From the above table, Z is maximum at point ‘B’ (2 , 6) i.e. The
Optimal Solution is:
X1 = 2 and
X2 = 6                                                                   ANSWER

                               http://www.rajeshtimane.com                        8
What is Feasibility ?
 Feasibility Region
  [Dictionary meaning of feasibility is possibility]

       “The region of acceptable values of the
         Decision Variables in relation to the
      given Constraints (and the Non-Negativity
                   Restrictions)”

                   http://www.rajeshtimane.com         9
What is an Optimal Solution ?

 It is the Feasible Solution which Optimizes.
  i.e. “provides the most beneficial result for the specified
  objective function”.

 Ex: If Objective function is Profit then Optimal Solution
  is the co-ordinate giving Maximum Value of „Z‟…
  While; if objective function is Cost then the optimum
  solution is the coordinate giving Minimum Value of „Z‟.

                     http://www.rajeshtimane.com              10
Convex Sets and LPP’s
 “If any two points are selected in the feasibility region
 and a line drawn through these points lies completely
 within this region, then this represents a Convex Set”.

       Convex Set                                 Non-convex Set


          A
                                                              A


                B                                         B




                    http://www.rajeshtimane.com                    11

Linear programming graphical method (feasibility)

  • 1.
  • 2.
    What is aMathematical Model ? F=ma ‘Mathematical Expressions’ o Here m and a are called as ‘Decision Variables’ o F can be called as ‘Objective Functions’ o Now, F can be controlled or restricted by limiting m or a … say m < 50 kg …here, m can be called as a ‘Constraint’ o Similarly if a > o …always, then this condition is called as ‘Non-Negativity Condition’ http://www.rajeshtimane.com 2
  • 3.
    Illustration: Maximize: Z =3x1 + 5x2 Objective Function Subject to restrictions: x1 <4 Functional 2x2 < 12 Constraints 3x1 + 2x2 < 18 Non negativity condition x1 >0 Non-negativity x2 >0 constraints http://www.rajeshtimane.com 3
  • 4.
    What is LinearProgramming (LP)?  The most common application of LP is allocating limited resources among competing activities in a best possible way i.e. the optimal way.  The adjective linear means that all the mathematical functions in this model are required to be linear functions.  The word programming does not refer to computer programming; rather, essentially a synonym for planning. http://www.rajeshtimane.com 4
  • 5.
    Graphical Solution Ex) Maximize:Z = 3x1 + 5x2 Subject to restrictions: x1 < 4 2x2 < 12 i.e. x2 < 6 3x1 + 2x2 < 18 Non negativity condition x1, x2 > 0 Solution: finding coordinates for the constraints (assuming perfect equality), by putting one decision variable equal to zero at a time. Restrictions (Constraints) Co-ordinates x1 < 4 (4 , 0) x2 < 6 (0 , 6) 3x1 + 2x2 < 18 (0 , 9) & (6 , 0) http://www.rajeshtimane.com 5
  • 6.
    Restrictions (Constraints) Co-ordinates Non-negativity Constraint x1 < 4 (4 , 0) x1, > 0 x2 < 6 (0 , 6) x2 > 0 3x1 + 2x2 < 18 (0 , 9) & (6 , 0) X2 10 8 A B 6 4 C Feasible Region (Shaded / Points A, B, C, D and E) 2 0 D E 2 4 6 8 10 X1 http://www.rajeshtimane.com 6
  • 7.
    Feasible Solutions  Tryco-ordinates of all the corner points of the feasible region.  The point which will lead to most satisfactory objective function will give Optimal Solution.  Note: for co-ordinates at intersection; solve the equations (constraints) of the two lines simultaneously. http://www.rajeshtimane.com 7
  • 8.
    Optimal Solution Corner Limiting Constraint Co-ordinate Max. Z= 3x1 + 5x2 A x2 = 6 (0 , 6) 30 B x2 = 6 & 3x1 + 2x2 = 18 (2 , 6) 36 C x1 = 4 & 3x1 + 2x2 = 18 (4 , 3) 27 D x1 = 4 (4 , 0) 12 E Origin (0 , 0) 0 From the above table, Z is maximum at point ‘B’ (2 , 6) i.e. The Optimal Solution is: X1 = 2 and X2 = 6 ANSWER http://www.rajeshtimane.com 8
  • 9.
    What is Feasibility?  Feasibility Region [Dictionary meaning of feasibility is possibility] “The region of acceptable values of the Decision Variables in relation to the given Constraints (and the Non-Negativity Restrictions)” http://www.rajeshtimane.com 9
  • 10.
    What is anOptimal Solution ?  It is the Feasible Solution which Optimizes. i.e. “provides the most beneficial result for the specified objective function”.  Ex: If Objective function is Profit then Optimal Solution is the co-ordinate giving Maximum Value of „Z‟… While; if objective function is Cost then the optimum solution is the coordinate giving Minimum Value of „Z‟. http://www.rajeshtimane.com 10
  • 11.
    Convex Sets andLPP’s “If any two points are selected in the feasibility region and a line drawn through these points lies completely within this region, then this represents a Convex Set”. Convex Set Non-convex Set A A B B http://www.rajeshtimane.com 11