OPERATIONS RESEARCH

CAT 1
21/05493 GICHERU DUNCAN
1. Under what conditions do LP problems have:
(a) Alternate (or multiple) optimal solution.
ANSWER
Linear programming (LP) problems can have alternate (or multiple) optimal
solutions under the following conditions:
1. Non-Unique Objective Coefficients: If the coefficients of the objective
function are not unique for determining the optimal solution, meaning multiple
combinations of decision variables result in the same objective function value,
then LP problems can have multiple optimal solutions.
2. Degeneracy: Degeneracy occurs when the number of binding constraints in
the solution is less than the number of decision variables. In such cases, there
can be multiple solutions that are all optimal. The presence of degeneracy is a
common reason for alternate optimal solutions.
3. Redundant Constraints: When there are redundant or linearly dependent
constraints in the LP problem, it can lead to multiple optimal solutions.
Redundant constraints do not affect the feasible region, so different
combinations of these constraints can result in different optimal solutions.

(b) Unbounded solution.

ANSWER
1. There is no feasible solution: If the feasible region (the region defined
by the constraints) is empty, meaning there are no values of x₁, x₂, ...,
xₙ that satisfy all the constraints, then the problem has an unbounded
solution. In this case, the objective function can be made arbitrarily
large or small without violating any constraints.

2. The feasible region is unbounded: Even if there are feasible solutions, the
problem can have unbounded solutions if the feasible region extends infinitely
in one or more directions. This means that there are certain values of the
decision variables that can be increased or decreased indefinitely while still
remaining within the feasible region. As a result, the objective function can be
increased or decreased without limit.
(c) Infeasible solution (No solution). [6 marks]
Answer
1. Incompatible Boundaries: In some cases, the lower and upper bounds on the
variables may be set in a way that makes it impossible to find a feasible
solution. For example, if you have a constraint 0 ≤ x ≤ -1, there is no real
number that can satisfy this constraint.
2. Empty Feasible Region: The feasible region of an LP problem is defined by
the intersection of the constraint boundaries. If this intersection is empty,
meaning there is no common area where all constraints overlap, the problem is
infeasible. This can occur when the constraints are positioned in a way that
they do not intersect.

2. Write the dual of the following Primal:

Max Z = -3x1 - x2
Subject to
- x1 - x2 ≤ -1
-2x1 - 3x2 ≤ - 2
x1 ≥ 0 , x2 ≥ 0 [4 marks]
ANSWER

Given Primal:
Max Z = -3x1 - x2 Subject to:
1. -x1 - x2 ≤ -1
2. -2x1 - 3x2 ≤ -2
3. x1 ≥ 0
4. x2 ≥ 0
Step 1: Write the dual variables for each constraint. Let y1 be the dual variable for
constraint (1), and y2 be the dual variable for constraint (2).
Step 2: Write the dual objective function. The dual objective function is to minimize
the sum of the products of the dual variables and the right-hand side values of the
primal constraints:
Min W = -1y1 - 2y2
Step 3: Formulate the dual linear program. Subject to:
1. y1 - 2*y2 ≥ -3 (dual of constraint 1)
2. y1 - 3*y2 ≥ -1 (dual of constraint 2)
3. y1 ≥ 0 (dual of constraint 3)
4. y2 ≥ 0 (dual of constraint 4)
So, the dual linear program is as follows:
Min W = -y1 - 2*y2 Subject to:
1. y1 - 2*y2 ≥ -3
2. y1 - 3*y2 ≥ -1
3. y1 ≥ 0
4. y2 ≥ 0

2. A and B are two products to be manufactured, unit profits are Tsh 40 and Tsh 35
respectively. Maximum materials available are 60 Kg and 96 man-hrs of labour. Each
unit of A needs 2 kg of materials and 3 man-hours, whereas each unit of B needs 4 kg
of materials and 3 man-hours. Using the graphical method find optimal level of A and
B to be manufactured. [10 marks]
ANSWER

 Let x be the number of units of product A to be manufactured.

 Let y be the number of units of product B to be manufactured.
Now, we can set up the objective function to maximize profit: Maximize Z = 40x +
35y
Subject to the following constraints:
1. Materials constraint: 2x + 4y ≤ 60 kg
2. Labor constraint: 3x + 3y ≤ 96 man-hours
3. Non-negativity constraints: x ≥ 0 and y ≥ 0
 When x = 0, y = 15.
 When y = 0, x = 30. Plot the line y = 15 - 0.5x.
Constraint 2 (Labor):
  When x = 0, y = 32.
 When y = 0, x = 32. Plot the line y = 32 - x.

3. Use the simplex method to solve the following LP problem.

M ax Z = 6x1 + 17x2 + 10x3
Subject to:
x1 + x2 + 4x3 ≤ 2000
2x1 + x2 + x3 ≤ 3600
x1 + 2x2 + 2x3 ≤ 2400
x1 ≤ 3044
and
x1, x2, x3 ≥ 0 [10 marks]
The standard form of an LP problem has the following characteristics:

All variables are non-negative.

All constraints are equality constraints.
The objective function is a minimization problem (maximization problems can be
converted by multiplying the objective function by -1).
In this case, your LP problem is already in standard form, as it's a maximization
problem with all variables non-negative and with inequality constraints. We'll start by
setting up the initial simplex table:

-------------------------------------
| Basic Variables | x1 | x2 | x3 | RHS |
-------------------------------------
| x1 + x2 + 4x3 | 1 | 1 | 4 | 2000 |
| 2x1 + x2 + x3 | 2 | 1 | 1 | 3600 |
| x1 + 2x2 + 2x3 | 1 | 2 | 2 | 2400 |
| -Z | -6 | -17 | -10 | 0 |
-------------------------------------
Now, let's use the simplex method to find the optimal solution:

Determine the pivot column: Select the most negative value in the bottom row (the
objective function row). In this case, it's -17, which corresponds to x2.

Determine the pivot row: Calculate the ratios of the right-hand side values to the
values in the pivot column and select the smallest positive ratio. In this case, we need
to choose the pivot row based on the ratios of RHS values to x2 coefficients.

For the first constraint: 2000 / 1 = 2000

For the second constraint: 3600 / 1 = 3600
For the third constraint: 2400 / 2 = 1200
The smallest positive ratio is 1200, which corresponds to the third constraint. So, the
pivot row is the third row.

Perform the pivot operation: Divide the pivot row by the pivot element (2 in this case)
to make the pivot element equal to 1. Then, use row operations to make all other
elements in the pivot column equal to 0.
The updated table becomes:
-------------------------------------
| Basic Variables | x1 | x2 | x3 | RHS |
-------------------------------------
| x1 + x2 + 4x3 | 0.5 | 0 | 2.5 | 1200 |
| 2x1 + x2 + x3 | 2 | 1 | 1 | 3600 |
| x1 + 2x2 + 2x3 | 0.5 | 1 | 1 | 1200 |
| -Z | -22 | 0 | -7 | 24000|
-------------------------------------
In this case, the bottom row has no negative values. The solution is optimal.
The optimal solution is Z = 24000, x1 = 1200, x2 = 0, and x3 = 0. These values
maximize the objective function Z.