1 Material to Cover  relationship between different types of models  incorrect to round real to integer variables  logical relationship: site selection.

Slides:



Advertisements
Similar presentations
Thursday, April 11 Some more applications of integer
Advertisements

Outline LP formulation of minimal cost flow problem
BU Decision Models Integer_LP1 Integer Optimization Summer 2013.
Linear Programming Problem. Introduction Linear Programming was developed by George B Dantzing in 1947 for solving military logistic operations.
Lesson 08 Linear Programming
Chapter 3 Workforce scheduling.
Linear Programming Problem
1 Lecture 2 Shortest-Path Problems Assignment Problems Transportation Problems.
BU BU Decision Models Networks 1 Networks Models Summer 2013.
Network Flows. 2 Ardavan Asef-Vaziri June-2013Transportation Problem and Related Topics Table of Contents Chapter 6 (Network Optimization Problems) Minimum-Cost.
Introduction to Algorithms
1 Lecture 3 MGMT 650 Sensitivity Analysis in LP Chapter 3.
Progress in Linear Programming Based Branch-and-Bound Algorithms
Lecture 5 Set Packing Problems Set Partitioning Problems
WOOD 492 MODELLING FOR DECISION SUPPORT Lecture 6 LP Assumptions.
McGraw-Hill/Irwin © The McGraw-Hill Companies, Inc., Integer Programming.
Integer Programming 3 Brief Review of Branch and Bound
Math443/543 Mathematical Modeling and Optimization
INTRODUCTION TO LINEAR PROGRAMMING
Review of Reservoir Problem OR753 October 29, 2014 Remote Sensing and GISc, IST.
Max-flow/min-cut theorem Theorem: For each network with one source and one sink, the maximum flow from the source to the destination is equal to the minimal.
Network Flows Based on the book: Introduction to Management Science. Hillier & Hillier. McGraw-Hill.
1 Lecture 4 Maximal Flow Problems Set Covering Problems.
1.3 Modeling with exponentially many constr.  Some strong formulations (or even formulation itself) may involve exponentially many constraints (cutting.
Chapter 7 Integer Programming Models
Network Models Tran Van Hoai Faculty of Computer Science & Engineering HCMC University of Technology Tran Van Hoai.
Network Optimization Problem
Linear Programming Topics General optimization model LP model and assumptions Manufacturing example Characteristics of solutions Sensitivity analysis Excel.
Spreadsheet Modeling & Decision Analysis:
Types of IP Models All-integer linear programs Mixed integer linear programs (MILP) Binary integer linear programs, mixed or all integer: some or all of.
Chapter 7 Transportation, Assignment & Transshipment Problems
EMIS 8373: Integer Programming “Easy” Integer Programming Problems: Network Flow Problems updated 11 February 2007.
Hub Location Problems Chapter 12
Operational Research & ManagementOperations Scheduling Workforce Scheduling 1.Days-Off Scheduling 2.Shift Scheduling 3. Cyclic Staffing Problem (& extensions)
Workforce scheduling – Days off scheduling 1. n is the max weekend demand n = max(n 1,n 7 ) Surplus number of employees in day j is u j = W – n j for.
Notes 5IE 3121 Knapsack Model Intuitive idea: what is the most valuable collection of items that can be fit into a backpack?
Lecture 6 – Integer Programming Models Topics General model Logic constraint Defining decision variables Continuous vs. integral solution Applications:
Chapter 1. Formulations 1. Integer Programming  Mixed Integer Optimization Problem (or (Linear) Mixed Integer Program, MIP) min c’x + d’y Ax +
DISTRIBUTION AND NETWORK MODELS (1/2)
Chap 10. Integer Prog. Formulations
© J. Christopher Beck Lecture 25: Workforce Scheduling 3.
EMIS 8374 Network Flow Models updated 29 January 2008.
Network Optimization Network optimization models: Special cases of linear programming models Important to identify problems that can be modeled as networks.
Lecture 5 – Integration of Network Flow Programming Models Topics Min-cost flow problem (general model) Mathematical formulation and problem characteristics.
Integer Programming (정수계획법)
IT Applications for Decision Making. Operations Research Initiated in England during the world war II Make scientifically based decisions regarding the.
1 An Arc-Path Model for OSPF Weight Setting Problem Dr.Jeffery Kennington Anusha Madhavan.
8/14/04 J. Bard and J. W. Barnes Operations Research Models and Methods Copyright All rights reserved Lecture 6 – Integer Programming Models Topics.
Integer Programming Definition of Integer Programming If requiring integer values is the only way in which a problem deviates from.
Lecture 6 – Integer Programming Models Topics General model Logic constraint Defining decision variables Continuous vs. integral solution Applications:
Introduction to Integer Programming Integer programming models Thursday, April 4 Handouts: Lecture Notes.
Approximation Algorithms Duality My T. UF.
Linear Programming Topics General optimization model
Chapter 1. Introduction Mathematical Programming (Optimization) Problem: min/max
Linear Programming Topics General optimization model
Lecture 5 – Integration of Network Flow Programming Models
1.3 Modeling with exponentially many constr.
Linear Programming Topics General optimization model
1.206J/16.77J/ESD.215J Airline Schedule Planning
Linear Programming Topics General optimization model
Integer Programming (정수계획법)
1.206J/16.77J/ESD.215J Airline Schedule Planning
Max Flow Application: Precedence Relations
Chapter 1. Formulations (BW)
1.3 Modeling with exponentially many constr.
Integer Programming (정수계획법)
Flow Feasibility Problems
Lecture 6 – Integer Programming Models
Chapter 1. Formulations.
REVIEW FOR EXAM 1 Chapters 3, 4, 5 & 6.
Presentation transcript:

1 Material to Cover  relationship between different types of models  incorrect to round real to integer variables  logical relationship: site selection  weak and strong formulation: uncapacitated facility location problem  set covering problems: airline crew scheduling  generalized piecewise linear approximation

2  max/min z = c 1 x 1 + c 2 x 2 + … + c n x n  s.t.  a i1 x 1 + a i2 x 2 + … + a in x n b i, i = 1,…, m  0  x j  u j, j = 1,…, n  x j integer for some or all j =1,…, n {    } Linear Integer Programming - IP

3  mixed IP (MIP): some x j  Z #1, some x j   #2  pure IP: all x j  Z  binary decision variable: x j = 1 or 0 (e.g., a variable for a yes-no decision)  binary IP (BIP): all x j being binary Linear Integer Programming - IP #1 Z: the set of integers; Z + : the set of positive integers #2  : the set of real numbers ;  + : the set of positive real numbers

4 Motivation of Studying IP  integer variables in some context  e.g., machine, manpower  logical relationship  incorrect to round continuous variables

5 Incorrect to Round Continuous Variables                     optimal LP solution X1X1 X2X2 optimal IP solution iso-cost line    usually all right to round in real life problems with large x i

6 Example to Motivate IP  site selection: three designs A, B, C on sites 1, 2, 3, 4  total amount for investment: $100 M  how to invest? OptionA1A1A2A2A3A3A4A4B1B1B2B2B3B3B4B4C1C1C2C2C3C3C4C4 Net Income ($M) Investment ($M)

7 Example to Motivate IP  I = {A, B, C, D}, J = {1, 2, 3, 4}  y ij = 1 iff design i used at site j, i  I, j  J  max z =  i  I  p ij y ij  s.t.  i  I  j  J a ij y ij  100  y ij  {0, 1}, i  I, j  J  optimal solution: y A1 = y A3 = y B3 = y B4 = y C1 = 1; z * = 40

8 Example to Motivate IP  boss says NO!  at most one design at a site  a building at site 2 (required)  at most two designs at the three sites  design A considered for sites 1, 2, and 3 only if being used at site 4  how to model?

9 Example to Motivate IP  at most one design at a site and a building at site 2 (required)  y A1 + y B1 + y C1  1, y A2 + y B2 + y C2 = 1,  y A3 + y B3 + y C3  1, y A4 + y B4 + y C4  1  design A considered for sites 1, 2, and 3 only if being used at site 4  y A1 + y A2 + y A3  3y A4

10 Example to Motivate IP  at most two designs at the three sites  w i = 1, if design i is used, = 0, o.w., i = A, B, C  w A + w B + w C  2  y i1 +y i2 +y i3 +y i4  4w i, i = A, B, C  optimal solution: y A1 = y A4 = y B2 = y B3 = 1; others = 0; z * = 37

11 Logical Constraints for Variables  n situations how to model (i) at most k of them hold, (ii) at least k of them hold, and (iii) exactly k hold  y j binary variables for j = 1 to n; y j = 1 if j holds, and = 0 otherwise  mutually exclusive y j : y 1 + y 2 + … + y n  1  at most k of y j = 1: y 1 + y 2 + … + y n  k  at least k of y j = 1: y 1 + y 2 + … + y n  k  exactly k of y j = 1: y 1 + y 2 + … + y n = k

12 Logical Constraints for Expressions  either-or constraints  either f 1 (x 1, …, x n )  b 1 or f 2 (x 1, …, x n )  b 2 or both  IP formulation: let y be a binary variable  M: a large positive number, practically “  ”  two constraints: f 1 (x 1, …, x n )  b 1 +My and f 2 (x 1, …, x n )  b 2 +M(1-y)  only one of these two being picked by the optimization procedure

13 Logical Constraints for Expressions  m constraints, at least k out of m being true  f 1 (x 1, …, x n )  b 1, …, f m (x 1, …, x n )  b m  modeling procedure  m binary variables y i, one for each constraint  f 1 (x 1, …, x n )  b 1 +M(1-y 1 ), …, f m (x 1, …, x n )  b m +M(1-y m )  y 1 + … + y m  k

14 An Example of Logical Constraints for Expressions  single processor for three jobs, of processing times 3 hr, 5 hr, and 7 hr, respectively  objective: minimizing the total completion time of the three jobs  how to formulate it as an integer program?  note. The IP is for the illustration of formulation. The problem has very simple solution.

15 Definitions of Parameters and Variables  s i : the processing start time of job i  c i : the completion time of job i  p i : the processing time of job i (i.e., p 1 = 3, p 2 = 5, p 3 = 7)  C:total completion time

16 How About This?  if job 1 before job 2, and before job 3, C = 26  if job 1 before job 3, and before job 2, C = 28  if job 2 before job 1, and before job 3, C = 28  if job 2 before job 3, and before job 1, C = 32  if job 3 before job 1, and before job 2, C = 32  if job 3 before job 2, and before job 1, C = 34

17 How About This?  s 1  s 2  s 3, C = 26  if (y 12 =1 & y 23 =1), C = 26  s 1  s 3  s 2, C = 28  if (y 13 =1 & y 32 =1), C = 26  s 2  s 1  s 3, C = 28  if (y 21 =1 & y 13 =1), C = 28  s 2  s 3  s 1, C = 32  if (y 23 =1 & y 31 =1), C = 32  s 3  s 1  s 2, C = 32  if (y 31 =1 & y 12 =1), C = 32  s 3  s 2  s 1, C = 34  if (y 32 =1 & y 21 =1), C = 34  then setting conditions on y ij …, which obviously not working

18 The Formulation  min c 1 + c 2 + c 3,  s.t. c 1 -s 1 = 3;c 2 -s 2 = 5;c 3 -s 3 = 7;  one of c 1  s 2 and c 2  s 1 holds;  one of c 1  s 3 and c 3  s 1 holds;  one of c 2  s 3 and c 3  s 2 holds;  c i  0, s i  0, i = 1, 2, 3.

19 The Formulation  min c 1 + c 2 + c 3,  s.t. c 1 -s 1 = 3;c 2 -s 2 = 5;c 3 -s 3 = 7;  c 1  s 2 +My 12 ; c 2  s 1 +My 21 ; y 12 +y 21 = 1;  c 1  s 3 +My 13 ; c 3  s 1 +My 31 ; y 13 +y 31 = 1;  c 2  s 3 +My 23 ; c 3  s 2 +My 32 ; y 23 +y 32 = 1;  c i  0, s i  0, i = 1, 2, 3;  y ij  {0, 1}, i  j, and i, j = 1, 2, 3 y ij = 1 if job j is before job i.

20 Equivalence Between BIP and General PIP  BIP  PIP  BIP  PIP  a PIP of bounded integer variables  BIP  max5x 1 + 2x 2  s.t.2x 1 + x 2  15  x 1  0, x 2  Z +  conversion  0  x 2  15  x 2 = y 1 + 2y 2 + 4y 3 + 8y 4, y i  binary

21 Fixed-Charge Problem  costs for having a facility at site j, j = 1 to n  set up cost k j  variable cost c j per unit of capacity  capacity of the whole system  C  minimum cost site selection for the capacity constraint

22 Fixed-Charge Problem n j=1 min  f j (x j ) where f j (x j ) = { k j + c j x j, if x j > 0 0, if x j = 0 k j = set-up cost, c j = per unit cost IP formulation: min n j =1  ( c j x j + k j y j ) s.t.s.t. x j  My j, j = 1, …, n;  j x j  C; y j  {0,1}, j = 1, …, n ; x j  0, j = 1, …, n ;

23 A More Realistic Fixed-Charge Problem  telecommunication network  source nodes S = {1, 3, 7}; destination nodes D = {2, 4, 5, 8}; transshipment node T = {6}  solid links: existing; dotted links: planned; total A = {1, 2, …, 17}  each link: (capacity, cost)  planned links A’ = {1, 2, …, 5}; fixed costs f 1 = 8; f 2 = 6; f 3 = 9, f 4 = 7, f 5 = 7  min cost construction to satisfy the demands & flows

24 A More Realistic Fixed-Charge Problem  decision variables  x k : the amount of flow in link k  y k : the binary variable of constructing link k  A'  parameters  b i : the demand of node i (source = -demand)  u k : the upper bound of flow of link k  f k :the fixed cost coefficient  c k :the variable cost coefficient

25 A More Realistic Fixed-Charge Problem  min z =  k  A’ f k y k +  k  A c k x k  s.t.  total in-flow – total out-flow = b i, conservation of flow  node i  x k  u k y k, capacity constraint  (proposed) arc k  A'  x k  u k, capacity constraint  (existing) arc k  A  y k = 0 or 1, binary variable  (proposed) arc k  A'  x k  0,  arc k  A  A'

26 Facility Location Problem  distributing goods to n customers possibly through m warehouses  warehouse i  fixed cost f i  variable cost v i per unit capacity  maximum capacity u i  shipment cost c ij per unit from warehouse i to customer j C1C1... Cn W1W1... Wm

27 Facility Location Problem  distributing goods to n customers possibly through m warehouses  data  d j : demand for customer j  u i : maximum capacity at warehouse i  f i : fixed cost to build warehouse i  v i : variable cost/unit of capacity of warehouse i  c ij : variable cost/unit of goods sent from warehouse i to customer j  decision variables  y i : build a warehouse at site i? (1 = yes, 0 = no)  z i : capacity (supply) of warehouse i  x ij : shipment from warehouse i to customer j

28 Facility Location Problem

29 Strong versus Weak Formulation – An Illustration through the Facility Location Problem  uncapacitated version of the facility location problem  intuitively optimal to have each customer satisfied by one warehouse  simplified the formulation  re-definition  c ij : shipment cost to customer j, possibly including the variable cost of operating warehouse i for demand d j  x ij : proportion of demand j satisfied by warehouse i

30 Weak Formulation

31 Strong Formulation

32 Comparison of the Strong and Weak Formulations  strong: more constraints  x ij  y i, i = 1, …, m; j = 1, …, n  weak: less constraints   i x ij  ny i, j = 1, …, n  which one is better?  strong: more precise constraints and possibly shorter computation time

33 Covering Problems and Partitioning Problems  S:a set of m items  S j :a subset of S that includes one or more of the items, j = 1, …, n  c j :the cost of selecting subset j  selecting the minimum cost collection of subsets S j to include elements of S  set covering: fine as long as including all items of S  set partitioning: each element of S is included exactly once

34 Airline Crew Scheduling (Set Covering Problem)  service network  group legs into tours according to constraints LAX SEA CHI DEN DFW

35 Airline Crew Scheduling (Set Covering Problem)  a tour: feasible assignment for a crew, starting & ending at DFW  a leg: a flight scheduled between two cities  covering 11 legs by 3 crews on 12 possible tours  minimizing the total tour cost

36 Airline Crew Scheduling (Set Covering Problem)  optimal solution: “Dead heading” on first leg Min2x2x 1 +3x3x 2 +4x4x 3 + … +8x8x 11 +9x9x 12

37 The Days-Off Scheduling Problem  (5,7)-cycle: 5 working days + 2 consecutive days off  7 days-off patterns  parameters  r i = number of employees required on day i  c j = weekly cost of pattern j per employee  decisions  x j = # of employees using days-off pattern j

38 The Days-Off Scheduling Problem  min z =  c j x j  s.t. (  x j ) – x i – x i-1  r i, i = 1,…7  x j  0 and integer, j = 1,…,7; x 0 = x 7 7 j=1 7 j=1  solve problem to get x * j  minimum cost workforce W =  x * j 7 j=1

39 The Days-Off Scheduling Problem  possibly to be solved as two LP  compact expression  Minimize z = cx  s.t. x  0 and integer

40 The Cutting Stock Problem  raw material: rolls of 25 ft  requirements  5-foot: 40 rolls  8-foot: 35 rolls  12-foot: 30 rolls  15-foot: 25 rolls  17-foot: 20 rolls  objective: using minimum # of 25-foot rolls

41 General Piecewise Linear Approximations  f j ( x j ), 0  x j  u j  r = number of grid points  (d ij, f ij ) be i th grid point, i = 1…, r

42 Linear Transformation for x j  x j =   i d ij  g j (x j ) =   i f ij    i = 1,  i  0, i = 1,…, r  not sufficient to guarantee that the solution is on one of the line segments r i=1 r i=1 r i=1

43 Additional Constraints for Piecewise Linear Approximation  condition: at most two positive  i, and positive  i ‘s adjacent  1 ≤ y 1  i ≤ y i -1 + y i, i = 2,…, r–1  r ≤ y r -1 y 1 + y 2 + · · · + y r -1 = 1 y i = 0 or 1, i = 1,..., r–1  not necessary to define  ’s if minimizing a convex function or maximizing a concave function

44 Approximation in Minimizing a Convex Objective Function y x1x1 CBAOpoints

45 Approximation in Minimizing a Convex Objective Function all right to omit (4) if approximating a convex objective function in minimization or a concave objective function in maximization

46 Approximation in Minimizing a Convex Objective Function all right to omit (4) if approximating a convex objective function in minimization or a concave objective function in maximization

47 Special Non-linear Objective Functions  machines: A to D  products: P, Q, R  potential sales: P  100, Q  100, R  100 prod mh time PQR available time A B C D

48 Special Non-linear Objective Functions  nonlinear unit profit from the products prod sales PQR  How to formulate the problem?

49 Special Non-linear Objective Functions  maxZ = f 1 (P) + f 2 (Q) + f 3 (R)  s.t.20P + 10Q + 10R  2400 (mh A)  12P + 28Q + 16R  2400 (mh B)  15P + 6Q + 16R  2400 (mh C)  10P + 15Q  2400 (mh D)  P  100, Q  100, R  100 (marketing)  P  0, Q  0, R  0  How to model f 1 (P), f 2 (Q), f 3 (R)

50 Special Non-linear Objective Functions  P i :# of sales of product P in the ith price range  Q i :# of sales of product Q in the ith price range  R i :# of sales of product R in the ith price range  object function: max  Z = 60P 1 +45P P 3 +40Q 1 +60Q Q 3 +20R 1 +70R 2 +20R 3

51 Special Non-linear Objective Functions  Z = 60P 1 +45P P 3 +40Q 1 +60Q Q 3 +20R 1 +70R 2 +20R 3 +  for P  0  P 1  30, 0  P 2  30, 0  P 3  40  concave prices, no additional constraints  for Q  0  Q 1  30, 0  Q 2  30, 0  Q 3  40  use second price segment only if Q 1 = 30  use third price segment only if Q 2 = 30  for R  0  R 1  30, 0  R 2  30, 0  R 3  40  use second and third price segments only if R 1 = 30

52 Special Non-linear Objective Functions  for Q  y Q2 = 1 if sales in segment 2 are made  = 0 otherwise  y Q3 = 1 if sales in segment 3 are made  = 0 otherwise  30y Q2  Q 1  30, 30y Q3  Q 2  30y Q2, 0  Q 3  40y Q3  for R  y R2 = 1 if sales in segments 2 or 3 are made  = 0 otherwise  30y R2  R 1  30, 0  R 2  30y R2, 0  R 3  40y R2

53 Other Examples  Traveling salesman problems  sequence dependent setup times  assembly line balancing