Integer Programming Optimisation Methods. 2 Lecture Outline 1Introduction 2Integer Programming 3Modeling with 0-1 (Binary) Variables 4Goal Programming.
Post on 20-Dec-2015
222 Views
Preview:
Transcript
Integer Programming
Optimisation Methods
2
Lecture Outline
1 Introduction
2 Integer Programming
3 Modeling with 0-1 (Binary) Variables
4 Goal Programming
5 Nonlinear Programming
3
Introduction
Integer programming is the extension of LP that solves problems requiring integer solutions.
Goal programming is the extension of LP that permits more than one objective to be stated.
Nonlinear programming is the case in which objectives or constraints are nonlinear.
All three above mathematical programming models are used when some of the basic assumptions of LP are made more or less restrictive.
4
Summary: Linear Programming Extensions
Integer Programming
Linear, integer solutions
Goal Programming
Linear, multiple objectives
Nonlinear Programming
Nonlinear objective and/or constraints
5
Integer Programming
Solution values must be whole numbers in integer programming .
There are three types of integer programs:
pure integer programming;
mixed-integer programming; and
0–1 integer programming.
6
Integer Programming
• The Pure Integer Programming problems are cases in which all variables are required to have integer values.
• The Mixed-Integer Programming problems are cases in which some, but not all, of the decision variables are required to have integer values.
• The Zero–One Integer Programming problems are special cases in which all the decision variables must have integer solution values of 0 or 1.
7
9.3 The Branch-and-Bound Method for Solving Pure Integer Programming Problems
In practice, most IPs are solved by some versions of the branch-and-bound procedure. Branch-and-bound methods implicitly enumerate all possible solutions to an IP.
By solving a single subproblem, many possible solutions may be eliminated from consideration.
Subproblems are generated by branching on an appropriately chosen fractional-valued variable.
8
Suppose that in a given subproblem (call it old subproblem), assumes a fractional value between the integers i and i+1. Then the two newly generated subproblems areNew Subproblem 1 Old subproblem + Constraint
New Subproblem 2 Old subproblem + Constraint
Key aspects of the branch-and-bound method for solving pure IPs If it is unnecessary to branch on a subproblem, we
say that it is fathomed.
ixi
1ixi
9
These three situations (for a max problem) result in a subproblem being fathomed The subproblem is infeasible, thus it cannot yield the optimal
solution to the IP.
The subproblem yield an optimal solution in which all variables have integer values. If this optimal solution has a better z-value than any previously obtained solution that is feasible in the IP, than it becomes a candidate solution, and its z-value becomes the current lower bound (LB) on the optimal z-value for the IP.
The optimal z-value for the subproblem does not exceed (in a max problem) the current LB, so it may be eliminated from consideration.
A subproblem may be eliminated from consideration in these situations The subproblem is infeasible.
The LB is at least as large as the z-value for the subproblem
10
Two general approaches are used to determine which subproblem should be solved next. The most widely used is LIFO.
LIFO leads us down one side of the branch-and-bound tree and quickly find a candidate solution and then we backtrack our way up to the top of the other side
The LIFO approach is often called backtracking.
The second commonly used approach is jumptracking. When branching on a node, the jumptracking method
solves all the problems created by branching.
11
When solving IP problems using Solver you can adjust a Solver tolerance setting.
The setting is found under the Options.
For example a tolerance value of .20 causes the Solver to stop when a feasible solution is found that has an objective function value within 20% of the optimal solution.
12
Harrison Electric Company
The Company produces two products popular with home renovators: old-fashioned chandeliers and ceiling fans.
Both the chandeliers and fans require a two-step production process involving wiring and assembly.
It takes about 2 hours to wire each chandelier and 3 hours to wire a ceiling fan. Final assembly of the chandeliers and fans requires 6 and 5 hours, respectively.
The production capability is such that only 12 hours of wiring time and 30 hours of assembly time are available.
13
If each chandelier produced nets the firm $7 and each fan $6, Harrison’s production mix decision can be formulated using LP as follows:
maximize profit = $7X1 + $6X2
subject to: 2X1 + 3X2 ≤ 12 (wiring hours)6X1 + 5X2 ≤ 30 (assembly hours) X1, X2 ≥ 0 (nonnegative)
X1 = number of chandeliers produced X2 = number of ceiling fans produced
Harrison Electric Company
14
With only two variables and two constraints, the graphical LP approach to generate the optimal solution is given below:
6X1 + 5X2 ≤ 30
+ = Possible Integer Solution
Optimal LP Solution(X1 = 33/4, X2 = 11/2, Profit = $35.25
2X1 + 3X2 ≤ 12
Harrison Electric Company
15
Integer Solution to Harrison Electric Co.
Optimal solution
Solution if rounding off
16
Rounding off is one way to reach integer solution values, but it often does not yield the best solution.
An important concept to understand is that an integer programming solution can never be better than the solution to the same LP problem.
The integer problem is usually worse in terms of higher cost or lower profit.
Integer Programming
17
Branch and Bound Method
Branch and Bound breaks the feasible solution region into sub-problems until an optimal solution is found.
There are Six Steps in Solving Integer Programming Maximization Problems by Branch and Bound.
The steps are given over the next several slides.
18
Branch and Bound Method: The Six Steps
1. Solve the original problem using LP.
If the answer satisfies the integer constraints, it is done.
If not, this value provides an initial upper bound.
2. Find any feasible solution that meets the integer constraints for use as a lower bound.
Usually, rounding down each variable will accomplish this.
19
Branch and Bound Method Steps:
3. Branch on one variable from Step 1 that does not have an integer value.
Split the problem into two sub-problems based on integer values that are immediately above and below the non-integer value.
For example, if X2 = 3.75 was in the final LP solution, introduce the constraint X2 ≥ 4 in the first sub-problem and X2 ≤ 3 in the second sub-problem.
4. Create nodes at the top of these new branches by solving the new problems.
20
Branch and Bound Method Steps:
5.
a) If a branch yields a solution to the LP problem that is not feasible, terminate the branch.
b) If a branch yields a solution to the LP problem that is feasible, but not an integer solution, go to step 6.
21
Branch and Bound Method Steps:
5. (continued)
c) If the branch yields a feasible integer solution, examine the value of the objective function.
If this value equals the upper bound, an optimal solution has been reached.
If it is not equal to the upper bound, but exceeds the lower bound, set it as the new lower bound and go to step 6.
Finally, if it is less than the lower bound, terminate this branch.
22
Branch and Bound Method Steps:
6. Examine both branches again and set the upper bound equal to the maximum value of the objective function at all final nodes.
If the upper bound equals the lower bound, stop.
If not, go back to step 3.
Minimization problems involve reversing the roles of the upper and lower bounds.
23
Harrison Electric Co:
Figure 11.1 shows graphically that the optimal, non-integer solution is
X1 = 3.75 chandeliersX2 = 1.5 ceiling fans
profit = $35.25 Since X1 and X2 are not integers, this solution is not
valid. The profit value of $35.25 will serve as an initial
upper bound. Note that rounding down gives X1 = 3, X2 = 1, profit =
$27, which is feasible and can be used as a lower bound.
24
Integer Solution: Creating Sub-problems
The problem is now divided into two sub-problems: A and B.
Consider branching on either variable that does not have an integer solution; pick X1 this time.
Subproblem Amaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≥ 4
Subproblem Bmaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≤ 3
25
Optimal Solution for Sub-problems
Optimal solutions are:Sub-problem A: X1 = 4; X2 = 1.2, profit=$35.20Sub-problem B: X1=3, X2=2, profit=$33.00(see figure on next slide)
Stop searching on the Subproblem B branch because it has an all-integer feasible solution. The $33 profit becomes the lower bound.
Subproblem A’s branch is searched further since it has a non-integer solution. The second upper bound becomes $35.20, replacing $35.25
from the first node.
26
Optimal Solution for Sub-problem
27
Sub-problems C and D
Subproblem A’s branching yields Subproblems C and D.
Subproblem Cmaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≥ 4
X2 ≥ 2Subproblem Dmaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≥ 4
X2 ≤ 1
28
Sub-problems C and D (continued)
Subproblem C has no feasible solution at all because the first two constraints are violated if the X1 ≥ 4 and X2 ≥ 2 constraints are observed. Terminate this branch and do not consider its
solution. Subproblem D’s optimal solution is
X1 = 4 , X2 = 1, profit = $35.16. This non-integer solution yields a new upper
bound of $35.16, replacing the original $35.20. Subproblems C and D, as well as the final branches
for the problem, are shown in the figure on the next slide.
29
Harrison Electric’s FullBranch and Bound Solution
30
Subproblems E and F
Finally, create subproblems E and F and solve for X1 and X2 with the added constraints X1 ≤ 4 and X1 ≥ 5. The subproblems and their solutions are:
Subproblem Emaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≥ 4 X1 ≤ 4
X2 ≤ 1Optimal solution for E: X1 = 4, X2 = 1, profit = $34
31
Subproblems E and F (continued)
Subproblem Fmaximize profit = $7X1 + $6X2
Subject to: 2X1 + 3X2 ≤ 12 6X1 + 5X2 ≤ 30 X1 ≥ 4 X1 ≥ 5
X2 ≤ 1Optimal solution for F: X1 = 5, X2 = 0, profit = $35
32
Goal Programming
Firms usually have more than one goal. For example,
maximizing total profit,
maximizing market share,
maintaining full employment,
providing quality ecological management,
minimizing noise level in the neighborhood, and
meeting numerous other non-economic goals. It is not possible for LP to have multiple goals unless they are all
measured in the same units (such as dollars),
a highly unusual situation. An important technique that has been developed to supplement LP is
called goal programming.
33
Goal Programming (continued)
Goal programming “satisfices,”
as opposed to LP, which tries to “optimize.”
Satisfice means coming as close as possible to reaching goals.
The objective function is the main difference between goal programming and LP.
In goal programming, the purpose is to minimize deviational variables,
which are the only terms in the objective function.
34
Example of Goal Programming Harrison Electric Revisited
Goals Harrison’s management wants to
achieve, each equal in priority:
Goal 1: to produce as much profit above $30 as possible during the production period.
Goal 2: to fully utilize the available wiring department hours.
Goal 3: to avoid overtime in the assembly department.
Goal 4: to meet a contract requirement to produce at least seven ceiling fans.
35
Example of Goal Programming Harrison Electric Revisited
Need a clear definition of deviational variables, such as :d1
– = underachievement of the profit target
d1+ = overachievement of the profit target
d2– = idle time in the wiring dept. (underused)
d2+ = overtime in the wiring dept. (overused)
d3– = idle time in the assembly dept. (underused)
d3+ = overtime in the wiring dept. (overused)
d4– = underachievement of the ceiling fan goal
d4+ = overachievement of the ceiling fan goal
36
Ranking Goals with Priority Levels
A key idea in goal programming is that one goal is more important than another. Priorities are assigned to each deviational variable.
A key idea in goal programming is that one goal is more important than another. Priorities are assigned to each deviational variable.
Priority 1 is infinitely more important than Priority 2, which is infinitely more important than the next goal, and so on.
Priority 1 is infinitely more important than Priority 2, which is infinitely more important than the next goal, and so on.
37
Analysis of First Goal
38
Analysis of First and Second Goals
39
Analysis of All Four Priority Goals
40
Goal Programming Versus Linear Programming
Multiple goals (instead of one goal)
Deviational variables minimized (instead of maximizing
profit or minimizing cost of LP)
“Satisficing” (instead of optimizing)
Deviational variables are real (and replace slack variables)
41
9.2 Formulating Integer Programming Problems
Practical solutions can be formulated as IPs.
The basics of formulating an IP model
42
Example 1: Capital Budgeting IP
Stockco is considering four investments
Each investment Yields a determined NPV
Requires a certain cash flow at the present time
Currently Stockco has $14,000 available for investment.
Formulate an IP whose solution will tell Stockco how to maximize the NPV obtained from the four investments.
43
Example 1: Solution
Begin by defining a variable for each decision that Stockco must make.
The NPV obtained by Stockco isTotal NPV obtained by Stocko = 16x1 + 22x2 + 12x3 +
8x4
Stockco’s objective function ismax z = 16x1 + 22x2 + 12x3 +8x4
Stockco faces the constraint that at most $14,000 can be invested.
Stockco’s 0-1 IP ismax z = 16x1 + 22x2 + 12x3 +8x4
s.t. 5x1 + 7x2 + 4x3 +3x4 ≤ 14xj = 0 or 1 (j = 1,2,3,4)
44
In a set-covering problem, each member of a given
set must be “covered” by an acceptable member of some set.
The objective of a set-covering problem is to minimize the number of elements in set 2 that are required to cover all the elements in set 1.
Set Covering as an IP
45
Piece-wise linear functions as IP
0-1 variables can be used to model optimization problems involving piecewise linear functions.
A piecewise linear function consists of several straight line segments.
The graph of the piecewise linear function is made of straight-line segments.
The points where the slope of the piecewise linear function changes are called the break points of the function.
A piecewise linear function is not a linear function so linear programming cannot be used to solve the optimization problem.
46
By using 0-1 variables, however, a piecewise linear function can be represented in a linear form. Suppose the piecewise linear function f (x) has break points .bbb n
,...,,21
Piece-wise linear functions as IP
)()1()()( may write
we,for linear is )( Because
)1(z x
:as written bemay ),10(number
somefor Then, ., someFor
1
1
1k
1
kkkk
kk
kkk
kk
kk
bfzbfzxf
bxbxf
bzb
xzz
bxbk
47
Suppose the piecewise linear function f (x) has break points . Step 1 Wherever f (x) occurs in the
optimization problem, replace f (x) by .
Step 2 Add the following constraints to the problem:
bzbzbz nnfff ...
2211
bbb n,...,,
21
ninior zyzzzyyy
yzyyzyyzyyzyz
ii
n
n
nnnnn
,...,2,1;1,...,2,1
1...
1...
,,...,,,
01021
121
11132321211
Piece-wise linear functions as IP
48
If a piecewise linear function f(x) involved in a formulation has the property that the slope of the f(x) becomes less favorable to the decision maker as x increases, then the tedious IP formulation is unnecessary.
LINDO can be used to solve pure and mixed IPs.
In addition to the optimal solution, the LINDO output also includes shadow prices and reduced costs.
LINGO and the Excel Solver can also be used to solve IPs.
Piece-wise linear functions as IP
49
9.4 The Branch-and-Bound Method for Solving Mixed Integer Programming Problems
In mixed IP, some variables are required to be integers and others are allowed to be either integer or non-integers.
To solve a mixed IP by the branch-and-bound method, modify the method by branching only on variables that are required to be integers.
For a solution to a subproblem to be a candidate solution, it need only assign integer values to those variables that are required to be integers
50
9.5 Solving Knapsack Problems by the Branch-and-Bound Method
A knapsack problem is an IP with a single constraint.
A knapsack problem in which each variable must be equal to 0 or 1 may be written as
When knapsack problems are solved by the branch-and-bound method, two aspects of the method greatly simplify. Due to each variable equaling 0 or 1, branching on xi
will yield in xi =0 and an xi =1 branch. The LP relaxation may be solved by inspection.
max z = c1x1 + c2x2 + ∙∙∙ + cnxn
s.t. a1x1 + a2x2 + ∙∙∙ + anxn ≤ b x1 = 0 or 1 (i = 1, 2, …, n)
51
9.6 Solving Combinatorial Optimization Problems by the Branch-and-Bound Method
A combinatorial optimization problem is any optimization problem that has a finite number of feasible solutions.
A branch-and-bound approach is often the most efficient way to solve them.
Examples of combinatorial optimization problems Ten jobs must be processed on a single machine. It is
known how long it takes to complete each job and the time at which each job must be completed. What ordering of the jobs minimizes the total delay of the 10 jobs?
52
A salesperson must visit each of the 10 cities before returning to his home. What ordering of the cities minimizes the total distance the salesperson must travel before returning home? This problem is called the traveling sales person problem (TSP).
In each of these problems, many possible solutions must be considered.
53
When using branch-and-bound methods to solve TSPs with many cities, large amounts of computer time is needed.
Heuristic methods, or heuristics, can be used to quickly lead to a good solution.
Heuristics is a method used to solve a problem by trial and error when an algorithm approach is impractical.
Two types of heuristic methods can be used to solve TSP; nearest neighbor method and cheapest-insertion method.
54
Nearest Neighbor Method Begin at any city and then “visit” the nearest city. Then go to the unvisited city closest to the city we have
most recently visited. Continue in this fashion until a tour is obtained. After
applying this procedure beginning at each city, take the best tour found.
Cheapest Insertion Method (CIM) Begin at any city and find its closest neighbor. Then create a subtour joining those two cities. Next, replace an arc in the subtour (say, arc (i, j) by the
combinations of two arcs---(i, k) and (k, j), where k is not in the current subtour---that will increase the length of the subtour by the smallest (or cheapest) amount.
Continue with this procedure until a tour is obtained. After applying this procedure beginning with each city, we take the best tour found.
55
Three methods to evaluate heuristics Performance guarantees
Gives a worse-case bound on how far away from optimality a tour constructed by the heuristic can be
Probabilistic analysis A heuristic is evaluated by assuming that the location
of cities follows some known probability distribution
Empirical analysis Heuristics are compared to the optimal solution for a
number of problems for which the optimal tour is known
56
• cij = distance from city i to city j
• xij = 1 if tour visits i then j, and 0 otherwise (binary)
• ti = arbitrary real numbers we need to solve for
n
i
n
jijij xc
1 1
njinnxtt
njx
nix
ijji
n
iij
n
jij
,21
11
11
0
0
minimize:
s.t.
An IP formulation can be used to solve a TSP but can become unwieldy and inefficient for large TSPs.
LINGO can be used to solve the IP of a TSP.
57
9.7 Implicit Enumeration
The method of implicit enumeration is often used to solve 0-1 IPs. Many IP problems can be converted to 0-1 IP problems.
Implicit enumeration uses the fact that each variable must be equal to 0 or 1 to simplify both the branching and bounding components of the branch-and-bound process and to determine efficiently when a node is infeasible.
The tree used in the implicit enumeration method is similar to those used to solve 0-1 knapsack problems.
Some nodes have variable that are specified called fixed variables.
58
All variables whose values are unspecified at a node are called free variables.
For any node, a specification of the values of all the free variables is called a completion of the node.
Three main ideas used in implicit enumeration Suppose we are at any node with fixed variables, is
there an easy way to find a good completion of the node that is feasible in the original 0-1 TSP?
Even if the best completion of a node is not feasible, the best completion gives us a bound on the best objective function value that can be obtained via feasible completion of the node. This bound can be used to eliminate a node from consideration.
59
At any node, is there an easy way to determine if all completions of the node are infeasible? In general, check whether a node has a feasible
completion by looking at each constraint and assigning each free variable the best value for satisfying the constraint.
60
9.8 Cutting Plane Algorithm
An alternative method to the branch-and-bound method is the cutting plane algorithm.
Summary of the cutting plane algorithm
Step 1 Find the optimal tableau for the IP’s programming relaxation. If all variables in the optimal solution assume integer values, we have found an optimal solution to the IP; otherwise, proceed to step2.
Step 2 Pick a constraint in the LP relaxation optimal tableau whose right-hand side has the fractional part closest to 1/2. This constraint will be used to generate a cut.
Step 2a For the constraint identified in step 2, write its right-hand side and each variable’s coefficient in the form [x]+ f, where 0 <= f < 1.
61
Step 2b Rewrite the constraint used to generate the cut as
All terms with integer coefficients = all terms with fractional coefficients
Then the cut is
All terms with fractional coefficients <= 0
Step 3 Use the simplex to find the optimal solution to the LP relaxation, with the cut as an additional constraint.
If all variables assume integer values in the optimal solution, we have found an optimal solution to the IP.
Otherwise, pick the constraint with the most fractional right-hand side and use it to generate another cut, which is added to the tableau.
We continue this process until we obtain a solution in which all variables are integers. This will be an optimal solution to the IP.
9.8 Cutting Plane Algorithm
62
A cut generated by the method above has the following properties: Any feasible point for the IP will satisfy the
cut
The current optimal solution to the LP relaxation will not satisfy the cut
9.8 Cutting Plane Algorithm
top related