SlideShare a Scribd company logo
1 of 17
Download to read offline
Lecture 5. Basic Feasible Solution

                                           Feng Chen

                  Department of Industrial Engineering and Logistics
                                    Management
                          Shanghai Jiao Tong University
                                   Mar 18, 2010



Mar 18 2010                                                                    1
                        ©Copyright Feng Chen 2004-2010. All rights reserved.
Enumerate Algorithm

                           min z  cT x
                           s.t. Ax  b
                                x0

              How to solve LP as above?

              If we can find all feasible solutions and the
              number of these feasible solutions is finite,
              then we can compare these solutions and
              select the optimal solution.


Mar 18 2010                                                                      Page 2
                          ©Copyright Feng Chen 2004-2010. All rights reserved.
Standard LP

                                                                •Convert methods
                                                                •“<=“ , Slack variables
                                         p                      •“>=“, excess variables
                  p’                                            •Free variable, x=x1-x2
              Standard LP                                       with x1 and x2 to be
                                                                nonnegative.


              Linear
              Programming                                    ? For standard LP, can we
                                                             get the extreme point by
                                                             algebraic calculations.




Mar 18 2010                                                                               Page 3
                       ©Copyright Feng Chen 2004-2010. All rights reserved.
Standard LP


               Any Linear Programming

                                    Converted

              Standard Linear Programming




Mar 18 2010                                                                   Page 4
                       ©Copyright Feng Chen 2004-2010. All rights reserved.
Basic Feasible Solution

   There must exists at least one optimal solution at
   extreme point.

 We need not to search all
 points, maybe good to consider
 these solutions only related to
 extreme points. For this goal,
 we introduce another concept
 of basic feasible solution.




Mar 18 2010                                                                Page 5
                    ©Copyright Feng Chen 2004-2010. All rights reserved.
Constraints Cases
Consider a system,     Ax  b               A  R m,n

• case 1: m=n                          • Case 3: m<n
  x1  x2  3                             x1  x2  3
  x1  2 x2  6                           x2  x3  1

• Case 2: m>n                      We like case 1 & 2, but
                                   no research. If research,
  x1  x2  3                      we must consider case 3.
  x1  2 x2  6                   This case will lead to
  x1  4 x2  5                    many feasible solutions.
                                   So, we will address this
                                   case.

Mar 18 2010                                                                 Page 6
                     ©Copyright Feng Chen 2004-2010. All rights reserved.
Which case is hard?
 Suppose we have standard constraints: Ax  b                                     A  R m,n
 Here m indicates the number of constraints, and n is the number of variables.
• Case 1: m=n       Such case is easy to get optimal solution, because there is
   x1  x2  3      always unique solution, we need not to do hard job to find the
   x1  2 x2  6   optimal one from many candidates (determined).

• Case 2: m>n       Such case seems easy, if we can have a method to determine its
                    infeasibility. (overdetermined)
 x1  x2  3
  x1  2 x2  6    We like case 1 & 2, it seems easy to deal with.
 x1  4 x2  5

• Case 3: m<n       The case seems hard to get optimal, we do not like, because
 x1  x2  3        they give me many solutions, and we need to find the best
  x 2  x3   1   one(undetermined).
                    This case will lead to many feasible solutions. So, we will address
                    such difficulty case.
    Mar 18 2010                                                                               Page 7
                           ©Copyright Feng Chen 2004-2010. All rights reserved.
Basic solution
•In case m<n
                       x1  x2  3
                       x2  x3  1
   How to find all feasible solutions? Because the number of variable
   is larger than the number of equation, we would like to fix some
   variables, and solve a normal equation system with left variables.

If we restrict x3 to a value, e.g., x3=0, we can obtain a
solution to the above system as (2,1,0);

 If x3=1, then the corresponding solution is (1,2,1);

 If x3=2, which leads to a solution (0,3,2).


Mar 18 2010                                                                    Page 8
                        ©Copyright Feng Chen 2004-2010. All rights reserved.
Nonbasic variables

Def. Given Ax=b with n variables and m constraints (n>m), a
solution called basic solution of the system is obtained by
•setting n-m variables equal to 0
•solving for the values of the remaining m variables



:nonbasic varibale                 :basic variable

 Def. If a basic solution is called basic feasible solution(bfs), if
 it is also feasible to the prime problem, generally to satisfy the
 sign restrictions.

Mar 18 2010                                                                  Page 9
                      ©Copyright Feng Chen 2004-2010. All rights reserved.
Basic Feasible Solution
                (Gia) max z  3x1  2 x2
                        s.t. 2 x1  x2  x3  100
                             x1  x2  x4  80
                              x1       x 5  40
                               x1 , x2 , x3 , x4 , x5  0


Step 1. Number of Variables =5; number of constraints=3;

   Step 1.1 How many of nonbasic variables?
                                                                               5-3=2
   Step 1.2 How many of basic variables?
                                                                               3
Step 2. Select x3, x4 as the nonbasic variables, and let x3= x4=0, we
get a basic solution of                          2 x1  x2  0  100
                       (20,60,0,0,20)                                          x1  x2  0  80
                                                                               x1     x 5  40
Mar 18 2010                                                                     x1 , x2 ,0,0, x5  0 Page 10
                        ©Copyright Feng Chen 2004-2010. All rights reserved.
Example

min z  4 x1  3 x2
                                     Basic                   Nonbasic                Basic
 s.t. x1  x2  y3  40              variables               Variables               Feasible
     2 x1  x2  y 4  60                                                            Solution
     x1 , x2 , y3 , y 4  0          x1, x2                  y3,y4                   (20,20,0,0)




Mar 18 2010                                                                                        Page 11
                              ©Copyright Feng Chen 2004-2010. All rights reserved.
min z  4 x1  3 x2
                                     Basic                    Nonbasic               Basic
 s.t. x1  x2  y3  40              variables                Variables              Feasible
     2 x1  x2  y 4  60                                                            Solution
     x1 , x2 , y3 , y 4  0          x1, x2                   y3,y4                  (20,20,0,0)
                                     x 1 , y3                 x2 ,y4                 (30,0,10,0)




Mar 18 2010                                                                                        Page 12
                              ©Copyright Feng Chen 2004-2010. All rights reserved.
Basic                    Nonbasic         Basic
 min z  4 x1  3 x2                       variables                Variables        Feasible
 s.t. x1  x2  y3  40                                                              Solution
      2 x1  x2  y 4  60                 x1, x2                   y3, y4           (20,20,0,0)
     x1 , x2 , y3 , y 4  0                x1 , y3                  x2 ,y4           (30,0,10,0)
                                           x1 ,y4                   x2, y3           (40,0,0,-20)
                                           x2 ,y3                   x1, y4           (0,60,-20,0)
                                           x2 ,y4                   x1 , y3          (0,40,0,20)
                                           y3 ,y4                   x1 , x2          (0,0,40,60)




Mar 18 2010                                                                                   Page 13
                              ©Copyright Feng Chen 2004-2010. All rights reserved.
BFS & Extreme Point


        BFS   Relationship                 Extreme Point




Mar 18 2010                                                            Page 14
                ©Copyright Feng Chen 2004-2010. All rights reserved.
BFS & Extreme Point
Theorem For any LP, there is a unique extreme point of the LP’s feasible
region corresponding to each basic feasible solution. Also, there is at least one
bfs corresponding to each extreme point of the feasible region.

Proof.
          Given a basic feasible solution
                             x=(x1,x2,…,xn-m, 0,0,…,0)


          The point related to the bfs is an extreme point.
          Assume x is not an extreme point, which means the
          there exist two different points , y,z

                     y                x
                                                              z                   Contradiction


Mar 18 2010                                                                                 Page 15
                           ©Copyright Feng Chen 2004-2010. All rights reserved.
Homework
  pp92. 1,2




Mar 18 2010                                                          Page 16
              ©Copyright Feng Chen 2004-2010. All rights reserved.
The End




Mar 18 2010                                                          17
              ©Copyright Feng Chen 2004-2010. All rights reserved.

More Related Content

What's hot

Saddlepoint approximations, likelihood asymptotics, and approximate condition...
Saddlepoint approximations, likelihood asymptotics, and approximate condition...Saddlepoint approximations, likelihood asymptotics, and approximate condition...
Saddlepoint approximations, likelihood asymptotics, and approximate condition...jaredtobin
 
Chapter 2 pertubation
Chapter 2 pertubationChapter 2 pertubation
Chapter 2 pertubationNBER
 
Chapter 3 projection
Chapter 3 projectionChapter 3 projection
Chapter 3 projectionNBER
 
Adomian decomposition method for solving higher order boundary value problems
Adomian decomposition method for solving higher order boundary value problemsAdomian decomposition method for solving higher order boundary value problems
Adomian decomposition method for solving higher order boundary value problemsAlexander Decker
 
Face Recognition Using Sign Only Correlation
Face Recognition Using Sign Only CorrelationFace Recognition Using Sign Only Correlation
Face Recognition Using Sign Only CorrelationIDES Editor
 
Initial-Population Bias in the Univariate Estimation of Distribution Algorithm
Initial-Population Bias in the Univariate Estimation of Distribution AlgorithmInitial-Population Bias in the Univariate Estimation of Distribution Algorithm
Initial-Population Bias in the Univariate Estimation of Distribution AlgorithmMartin Pelikan
 
Six sigma quick references
Six sigma quick referencesSix sigma quick references
Six sigma quick referencesVIVOCORP
 
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片Chyi-Tsong Chen
 
Solution of linear and nonlinear partial differential equations using mixture...
Solution of linear and nonlinear partial differential equations using mixture...Solution of linear and nonlinear partial differential equations using mixture...
Solution of linear and nonlinear partial differential equations using mixture...Alexander Decker
 
11.solution of linear and nonlinear partial differential equations using mixt...
11.solution of linear and nonlinear partial differential equations using mixt...11.solution of linear and nonlinear partial differential equations using mixt...
11.solution of linear and nonlinear partial differential equations using mixt...Alexander Decker
 
Hash Functions: lecture series by Ahto Buldas
Hash Functions: lecture series by Ahto BuldasHash Functions: lecture series by Ahto Buldas
Hash Functions: lecture series by Ahto BuldasGuardTimeEstonia
 
Optimalpolicyhandout
OptimalpolicyhandoutOptimalpolicyhandout
OptimalpolicyhandoutNBER
 
Chapter 1 nonlinear
Chapter 1 nonlinearChapter 1 nonlinear
Chapter 1 nonlinearNBER
 
Chapter 5 heterogeneous
Chapter 5 heterogeneousChapter 5 heterogeneous
Chapter 5 heterogeneousNBER
 
Hedging, Arbitrage, and Optimality with Superlinear Frictions
Hedging, Arbitrage, and Optimality with Superlinear FrictionsHedging, Arbitrage, and Optimality with Superlinear Frictions
Hedging, Arbitrage, and Optimality with Superlinear Frictionsguasoni
 

What's hot (16)

Saddlepoint approximations, likelihood asymptotics, and approximate condition...
Saddlepoint approximations, likelihood asymptotics, and approximate condition...Saddlepoint approximations, likelihood asymptotics, and approximate condition...
Saddlepoint approximations, likelihood asymptotics, and approximate condition...
 
Chapter 2 pertubation
Chapter 2 pertubationChapter 2 pertubation
Chapter 2 pertubation
 
Chapter 3 projection
Chapter 3 projectionChapter 3 projection
Chapter 3 projection
 
Adomian decomposition method for solving higher order boundary value problems
Adomian decomposition method for solving higher order boundary value problemsAdomian decomposition method for solving higher order boundary value problems
Adomian decomposition method for solving higher order boundary value problems
 
Face Recognition Using Sign Only Correlation
Face Recognition Using Sign Only CorrelationFace Recognition Using Sign Only Correlation
Face Recognition Using Sign Only Correlation
 
Initial-Population Bias in the Univariate Estimation of Distribution Algorithm
Initial-Population Bias in the Univariate Estimation of Distribution AlgorithmInitial-Population Bias in the Univariate Estimation of Distribution Algorithm
Initial-Population Bias in the Univariate Estimation of Distribution Algorithm
 
Six sigma quick references
Six sigma quick referencesSix sigma quick references
Six sigma quick references
 
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片
Reliability-Based Design Optimization Using a Cell Evolution Method ~陳奇中教授演講投影片
 
Solution of linear and nonlinear partial differential equations using mixture...
Solution of linear and nonlinear partial differential equations using mixture...Solution of linear and nonlinear partial differential equations using mixture...
Solution of linear and nonlinear partial differential equations using mixture...
 
11.solution of linear and nonlinear partial differential equations using mixt...
11.solution of linear and nonlinear partial differential equations using mixt...11.solution of linear and nonlinear partial differential equations using mixt...
11.solution of linear and nonlinear partial differential equations using mixt...
 
Hash Functions: lecture series by Ahto Buldas
Hash Functions: lecture series by Ahto BuldasHash Functions: lecture series by Ahto Buldas
Hash Functions: lecture series by Ahto Buldas
 
Optimalpolicyhandout
OptimalpolicyhandoutOptimalpolicyhandout
Optimalpolicyhandout
 
Chapter 1 nonlinear
Chapter 1 nonlinearChapter 1 nonlinear
Chapter 1 nonlinear
 
Chapter 5 heterogeneous
Chapter 5 heterogeneousChapter 5 heterogeneous
Chapter 5 heterogeneous
 
Holographic Cotton Tensor
Holographic Cotton TensorHolographic Cotton Tensor
Holographic Cotton Tensor
 
Hedging, Arbitrage, and Optimality with Superlinear Frictions
Hedging, Arbitrage, and Optimality with Superlinear FrictionsHedging, Arbitrage, and Optimality with Superlinear Frictions
Hedging, Arbitrage, and Optimality with Superlinear Frictions
 

Similar to Or lpp

02.03 Artificial Intelligence: Search by Optimization
02.03 Artificial Intelligence: Search by Optimization02.03 Artificial Intelligence: Search by Optimization
02.03 Artificial Intelligence: Search by OptimizationAndres Mendez-Vazquez
 
Linearprog, Reading Materials for Operational Research
Linearprog, Reading Materials for Operational Research Linearprog, Reading Materials for Operational Research
Linearprog, Reading Materials for Operational Research Derbew Tesfa
 
Balaji-opt-lecture5-linear program sp13.ppt
Balaji-opt-lecture5-linear program sp13.pptBalaji-opt-lecture5-linear program sp13.ppt
Balaji-opt-lecture5-linear program sp13.pptgokulkumaraguru8
 
Balaji-opt-lecture6-act.ppt
Balaji-opt-lecture6-act.pptBalaji-opt-lecture6-act.ppt
Balaji-opt-lecture6-act.pptJamesGreen666883
 
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Matthew Leingang
 
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Mel Anthony Pepito
 
Lesson 27: Integration by Substitution (Section 4 version)
Lesson 27: Integration by Substitution (Section 4 version)Lesson 27: Integration by Substitution (Section 4 version)
Lesson 27: Integration by Substitution (Section 4 version)Matthew Leingang
 
The Use of Fuzzy Optimization Methods for Radiation.ppt
The Use of Fuzzy Optimization Methods for Radiation.pptThe Use of Fuzzy Optimization Methods for Radiation.ppt
The Use of Fuzzy Optimization Methods for Radiation.ppteslameslam18
 
primal and dual problem
primal and dual problemprimal and dual problem
primal and dual problemYash Lad
 
Integer Programming PPt.ernxzamnbmbmspdf
Integer Programming PPt.ernxzamnbmbmspdfInteger Programming PPt.ernxzamnbmbmspdf
Integer Programming PPt.ernxzamnbmbmspdfRaja Manyam
 
Lesson 25: Indeterminate Forms and L'Hôpital's Rule
Lesson 25: Indeterminate Forms and L'Hôpital's RuleLesson 25: Indeterminate Forms and L'Hôpital's Rule
Lesson 25: Indeterminate Forms and L'Hôpital's RuleMatthew Leingang
 
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...jensenbo
 
Duality in Linear Programming
Duality in Linear ProgrammingDuality in Linear Programming
Duality in Linear Programmingjyothimonc
 
Lesson 27: Integration by Substitution (Section 10 version)
Lesson 27: Integration by Substitution (Section 10 version)Lesson 27: Integration by Substitution (Section 10 version)
Lesson 27: Integration by Substitution (Section 10 version)Matthew Leingang
 
Lecture5-7_12946_Linear Programming The Graphical Method.pptx
Lecture5-7_12946_Linear Programming The Graphical Method.pptxLecture5-7_12946_Linear Programming The Graphical Method.pptx
Lecture5-7_12946_Linear Programming The Graphical Method.pptxhlKh4
 

Similar to Or lpp (20)

Unit.4.integer programming
Unit.4.integer programmingUnit.4.integer programming
Unit.4.integer programming
 
02.03 Artificial Intelligence: Search by Optimization
02.03 Artificial Intelligence: Search by Optimization02.03 Artificial Intelligence: Search by Optimization
02.03 Artificial Intelligence: Search by Optimization
 
Interview Preparation
Interview PreparationInterview Preparation
Interview Preparation
 
Linearprog, Reading Materials for Operational Research
Linearprog, Reading Materials for Operational Research Linearprog, Reading Materials for Operational Research
Linearprog, Reading Materials for Operational Research
 
Midterm I Review
Midterm I ReviewMidterm I Review
Midterm I Review
 
Balaji-opt-lecture5-linear program sp13.ppt
Balaji-opt-lecture5-linear program sp13.pptBalaji-opt-lecture5-linear program sp13.ppt
Balaji-opt-lecture5-linear program sp13.ppt
 
Balaji-opt-lecture6-act.ppt
Balaji-opt-lecture6-act.pptBalaji-opt-lecture6-act.ppt
Balaji-opt-lecture6-act.ppt
 
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
 
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
Lesson 17: Indeterminate forms and l'Hôpital's Rule (slides)
 
Lesson 27: Integration by Substitution (Section 4 version)
Lesson 27: Integration by Substitution (Section 4 version)Lesson 27: Integration by Substitution (Section 4 version)
Lesson 27: Integration by Substitution (Section 4 version)
 
The Use of Fuzzy Optimization Methods for Radiation.ppt
The Use of Fuzzy Optimization Methods for Radiation.pptThe Use of Fuzzy Optimization Methods for Radiation.ppt
The Use of Fuzzy Optimization Methods for Radiation.ppt
 
primal and dual problem
primal and dual problemprimal and dual problem
primal and dual problem
 
Integer Programming PPt.ernxzamnbmbmspdf
Integer Programming PPt.ernxzamnbmbmspdfInteger Programming PPt.ernxzamnbmbmspdf
Integer Programming PPt.ernxzamnbmbmspdf
 
Lesson 25: Indeterminate Forms and L'Hôpital's Rule
Lesson 25: Indeterminate Forms and L'Hôpital's RuleLesson 25: Indeterminate Forms and L'Hôpital's Rule
Lesson 25: Indeterminate Forms and L'Hôpital's Rule
 
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...
2008 : A Case Study: How to Speed Up the Simplex Algorithms on Problems Minim...
 
Duality in Linear Programming
Duality in Linear ProgrammingDuality in Linear Programming
Duality in Linear Programming
 
Lesson 27: Integration by Substitution (Section 10 version)
Lesson 27: Integration by Substitution (Section 10 version)Lesson 27: Integration by Substitution (Section 10 version)
Lesson 27: Integration by Substitution (Section 10 version)
 
Unit 4
Unit 4Unit 4
Unit 4
 
Unit 4
Unit 4Unit 4
Unit 4
 
Lecture5-7_12946_Linear Programming The Graphical Method.pptx
Lecture5-7_12946_Linear Programming The Graphical Method.pptxLecture5-7_12946_Linear Programming The Graphical Method.pptx
Lecture5-7_12946_Linear Programming The Graphical Method.pptx
 

Or lpp

  • 1. Lecture 5. Basic Feasible Solution Feng Chen Department of Industrial Engineering and Logistics Management Shanghai Jiao Tong University Mar 18, 2010 Mar 18 2010 1 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 2. Enumerate Algorithm min z  cT x s.t. Ax  b x0 How to solve LP as above? If we can find all feasible solutions and the number of these feasible solutions is finite, then we can compare these solutions and select the optimal solution. Mar 18 2010 Page 2 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 3. Standard LP •Convert methods •“<=“ , Slack variables p •“>=“, excess variables p’ •Free variable, x=x1-x2 Standard LP with x1 and x2 to be nonnegative. Linear Programming ? For standard LP, can we get the extreme point by algebraic calculations. Mar 18 2010 Page 3 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 4. Standard LP Any Linear Programming Converted Standard Linear Programming Mar 18 2010 Page 4 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 5. Basic Feasible Solution There must exists at least one optimal solution at extreme point. We need not to search all points, maybe good to consider these solutions only related to extreme points. For this goal, we introduce another concept of basic feasible solution. Mar 18 2010 Page 5 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 6. Constraints Cases Consider a system, Ax  b A  R m,n • case 1: m=n • Case 3: m<n x1  x2  3 x1  x2  3  x1  2 x2  6  x2  x3  1 • Case 2: m>n We like case 1 & 2, but no research. If research, x1  x2  3 we must consider case 3.  x1  2 x2  6 This case will lead to x1  4 x2  5 many feasible solutions. So, we will address this case. Mar 18 2010 Page 6 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 7. Which case is hard? Suppose we have standard constraints: Ax  b A  R m,n Here m indicates the number of constraints, and n is the number of variables. • Case 1: m=n Such case is easy to get optimal solution, because there is x1  x2  3 always unique solution, we need not to do hard job to find the  x1  2 x2  6 optimal one from many candidates (determined). • Case 2: m>n Such case seems easy, if we can have a method to determine its infeasibility. (overdetermined) x1  x2  3  x1  2 x2  6 We like case 1 & 2, it seems easy to deal with. x1  4 x2  5 • Case 3: m<n The case seems hard to get optimal, we do not like, because x1  x2  3 they give me many solutions, and we need to find the best  x 2  x3   1 one(undetermined). This case will lead to many feasible solutions. So, we will address such difficulty case. Mar 18 2010 Page 7 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 8. Basic solution •In case m<n x1  x2  3  x2  x3  1 How to find all feasible solutions? Because the number of variable is larger than the number of equation, we would like to fix some variables, and solve a normal equation system with left variables. If we restrict x3 to a value, e.g., x3=0, we can obtain a solution to the above system as (2,1,0);  If x3=1, then the corresponding solution is (1,2,1);  If x3=2, which leads to a solution (0,3,2). Mar 18 2010 Page 8 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 9. Nonbasic variables Def. Given Ax=b with n variables and m constraints (n>m), a solution called basic solution of the system is obtained by •setting n-m variables equal to 0 •solving for the values of the remaining m variables :nonbasic varibale :basic variable Def. If a basic solution is called basic feasible solution(bfs), if it is also feasible to the prime problem, generally to satisfy the sign restrictions. Mar 18 2010 Page 9 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 10. Basic Feasible Solution (Gia) max z  3x1  2 x2 s.t. 2 x1  x2  x3  100 x1  x2  x4  80 x1  x 5  40 x1 , x2 , x3 , x4 , x5  0 Step 1. Number of Variables =5; number of constraints=3; Step 1.1 How many of nonbasic variables? 5-3=2 Step 1.2 How many of basic variables? 3 Step 2. Select x3, x4 as the nonbasic variables, and let x3= x4=0, we get a basic solution of 2 x1  x2  0  100 (20,60,0,0,20) x1  x2  0  80 x1  x 5  40 Mar 18 2010 x1 , x2 ,0,0, x5  0 Page 10 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 11. Example min z  4 x1  3 x2 Basic Nonbasic Basic s.t. x1  x2  y3  40 variables Variables Feasible 2 x1  x2  y 4  60 Solution x1 , x2 , y3 , y 4  0 x1, x2 y3,y4 (20,20,0,0) Mar 18 2010 Page 11 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 12. min z  4 x1  3 x2 Basic Nonbasic Basic s.t. x1  x2  y3  40 variables Variables Feasible 2 x1  x2  y 4  60 Solution x1 , x2 , y3 , y 4  0 x1, x2 y3,y4 (20,20,0,0) x 1 , y3 x2 ,y4 (30,0,10,0) Mar 18 2010 Page 12 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 13. Basic Nonbasic Basic min z  4 x1  3 x2 variables Variables Feasible s.t. x1  x2  y3  40 Solution 2 x1  x2  y 4  60 x1, x2 y3, y4 (20,20,0,0) x1 , x2 , y3 , y 4  0 x1 , y3 x2 ,y4 (30,0,10,0) x1 ,y4 x2, y3 (40,0,0,-20) x2 ,y3 x1, y4 (0,60,-20,0) x2 ,y4 x1 , y3 (0,40,0,20) y3 ,y4 x1 , x2 (0,0,40,60) Mar 18 2010 Page 13 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 14. BFS & Extreme Point BFS Relationship Extreme Point Mar 18 2010 Page 14 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 15. BFS & Extreme Point Theorem For any LP, there is a unique extreme point of the LP’s feasible region corresponding to each basic feasible solution. Also, there is at least one bfs corresponding to each extreme point of the feasible region. Proof. Given a basic feasible solution x=(x1,x2,…,xn-m, 0,0,…,0) The point related to the bfs is an extreme point. Assume x is not an extreme point, which means the there exist two different points , y,z y x z Contradiction Mar 18 2010 Page 15 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 16. Homework pp92. 1,2 Mar 18 2010 Page 16 ©Copyright Feng Chen 2004-2010. All rights reserved.
  • 17. The End Mar 18 2010 17 ©Copyright Feng Chen 2004-2010. All rights reserved.