Scribd
Upload
38 views121 pages

Optimization - Integer Programming2

The document discusses integer programming and mixed integer linear programming (MILP), highlighting methods such as the Simplex Algorithm and Branch-and-Bound for solving optimization problems. It emphasizes the importance of understanding and decision-making in production planning and provides algorithms for finding optimal solutions. Key concepts include relaxing integer constraints, solving linear programming problems, and utilizing domain knowledge in the branching process.

Uploaded by

jf923123
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
38 views121 pages

Optimization - Integer Programming2

The document discusses integer programming and mixed integer linear programming (MILP), highlighting methods such as the Simplex Algorithm and Branch-and-Bound for solving optimization problems. It emphasizes the importance of understanding and decision-making in production planning and provides algorithms for finding optimal solutions. Key concepts include relaxing integer constraints, solving linear programming problems, and utilizing domain knowledge in the branching process.

Uploaded by

jf923123
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 121

Optimization – Integer

programming
Siva
Optional: Gomory cuts and Job shop scheduling(81-100)
Integer programming and mixed
integer linear programming(MILP)
Two things in life:
Understanding what is happening!!
Deciding what to do!!
An example: Production planning
Integer programming – One time linear
programming and rounding off
i) Relax the integer constraints to real numbers.
ii) Solve the Linear programming problem.
iii) Round off the solution to the nearest integers.
( Profit from x1 is 6 per unit and from x2 it is 8.)
Simplex Algorithm: Explanation with Example

The Simplex Algorithm is an iterative method used to solve


Linear Programming (LP) problems. It is particularly efficient
for solving optimization problems involving a linear objective
function subject to linear equality and inequality constraints.
Very Important problem
Branch-and-Bound Example:
Knapsack Problem
Modelling with Mixed Integer
Linear Programming
Branch and bound using domain
knowledge – Direct approach
This solution is feasible, but not optimal.
Integer programming – Solution by
branch and bound
Algorithm BranchAndBound:
Input: Problem instance
Output: Optimal solution (if exists)

1. Initialize:
- Create an empty list of subproblems (active set). d. If the bound suggests potential for better
- Add the root problem (entire solution space) to the active solutions:
set. - Branch the subproblem into smaller
- Set the best-known solution to null (or an infeasible large subproblems by partitioning the feasible region.
value for minimization). - Add the new subproblems to the active set.

2. While the active set is not empty: 3. Return the best-known solution.
a. Select a subproblem from the active set (e.g., using
depth-first, breadth-first, or best-first search).
b. Remove the selected subproblem from the active set.

c. Solve the relaxation of the subproblem to compute a


bound:
- If the bound is worse than the current best-known
solution, discard the subproblem.
- If the bound improves the best-known solution and
satisfies all constraints:
- Update the best-known solution.
Algorithm BranchAndBound_Minimization: c. Bounding:
Input: Integer programming problem (minimize c^T x - If z* ≥ best_objective, discard P.
subject to Ax ≤ b, x ∈ Z^n) - If x* is integer and feasible:
Output: Optimal integer solution (if exists) If z* < best_objective:
best_objective ← z*
1. Initialize: best_solution ← x*
- Create an empty list of subproblems (active set).
- Add the root problem (original LP relaxation) to the
active set.
- Set best_solution ← null, best_objective ← +∞.

2. While the active set is not empty:


a. Select and remove a subproblem P from the active
set.

b. Solve the LP relaxation of P:


- Let x* be the solution and z* be the objective
value.

You might also like