Line 17: Line 17:
 
----
 
----
  
'''Theorem: '''
+
'''Theorem: '''  
  
A basic feasible solution is optimal if and only if the corresponding reduced cost coefficeints are all nonnegative.
+
A basic feasible solution is optimal if and only if the corresponding reduced cost coefficeints are all nonnegative.  
  
 
'''Simplex Method:'''  
 
'''Simplex Method:'''  
Line 25: Line 25:
 
1. Transform the given problem into standard form by introducing slack variables&nbsp;<span class="texhtml">''x''<sub>3</sub></span>&nbsp;and&nbsp;<span class="texhtml">''x''<sub>4</sub></span>.<span class="texhtml"><sub></sub></span>  
 
1. Transform the given problem into standard form by introducing slack variables&nbsp;<span class="texhtml">''x''<sub>3</sub></span>&nbsp;and&nbsp;<span class="texhtml">''x''<sub>4</sub></span>.<span class="texhtml"><sub></sub></span>  
  
2. Form a canonical augmented matrix corresponding to an initial basic feasible solution.
+
2. Form a canonical augmented matrix corresponding to an initial basic feasible solution.  
  
3. Calculate the reduced cost coefficients corresponding to the nonbasic variables.
+
3. Calculate the reduced cost coefficients corresponding to the nonbasic variables.  
  
4. If&nbsp;<math>r_{j}>\geq0</math>&nbsp;for all&nbsp;<math>j</math>, stop. -- the current basic feasible solution is optimal.
+
4. If&nbsp;<math>r_{j}>\geq0</math>&nbsp;for all&nbsp;<span class="texhtml">''j''</span>, stop. -- the current basic feasible solution is optimal.  
  
5. Select a&nbsp;<math>q</math>&nbsp;such that&nbsp;<math>r_{q}<0</math>
+
5. Select a&nbsp;<span class="texhtml">''q''</span>&nbsp;such that&nbsp;<span class="texhtml">''r''<sub>''q''</sub> &lt; 0</span>  
  
6. If no&nbsp;<math>y_{iq}>0</math>, stop. -- the problem is unbounded; else, calculate&nbsp;<math>p=argmin_{i}\left \{ y_{i0}/y_{iq}:y_{iq}>0 \right \}</math>.&nbsp;
+
6. If no&nbsp;<span class="texhtml">''y''<sub>''i''''q''</sub> &gt; 0</span>, stop. -- the problem is unbounded; else, calculate&nbsp;<math>p=argmin_{i}\left \{ y_{i0}/y_{iq}:y_{iq}>0 \right \}</math>.&nbsp;  
  
7. Update the canonical augmented matrix by pivoting about the&nbsp;<math>(p, q)</math>&nbsp;th element.
+
7. Update the canonical augmented matrix by pivoting about the&nbsp;<span class="texhtml">(''p'',''q'')</span>&nbsp;th element.  
  
 
8. Go to step 3.<br>  
 
8. Go to step 3.<br>  
Line 113: Line 113:
 
Go to  
 
Go to  
  
*Part 1: [[ECE-QE AC3-2011 solusion-1|solutions and discussions]]<br>
+
*Part 1: [[ECE-QE AC3-2011 solusion-1|solutions and discussions]]
 +
*Part 2: [[ECE-QE AC3-2011 solusion-2|solutions and discussions]]
 
*Part 3: [[ECE-QE AC3-2011 solusion-3|solutions and discussions]]  
 
*Part 3: [[ECE-QE AC3-2011 solusion-3|solutions and discussions]]  
 
*Part 4: [[ECE-QE AC3-2011 solusion-4|solutions and discussions]]  
 
*Part 4: [[ECE-QE AC3-2011 solusion-4|solutions and discussions]]  

Revision as of 20:05, 28 June 2012

ECE Ph.D. Qualifying Exam in "Automatic Control" (AC)

Question 3, Part 2, August 2011

Part 1,2,3,4,5

 $ \color{blue}\text{2. } \left( \text{20 pts} \right) \text{ Use the simplex method to solve the problem, } $

               maximize        x1 + x2

               $ \text{subject to } x_{1}-x_{2}\leq2 $
                                        $ x_{1}+x_{2}\leq6 $                                         

                                        $ x_{1},-x_{2}\geq0. $


Theorem: 

A basic feasible solution is optimal if and only if the corresponding reduced cost coefficeints are all nonnegative.

Simplex Method:

1. Transform the given problem into standard form by introducing slack variables x3 and x4.

2. Form a canonical augmented matrix corresponding to an initial basic feasible solution.

3. Calculate the reduced cost coefficients corresponding to the nonbasic variables.

4. If $ r_{j}>\geq0 $ for all j, stop. -- the current basic feasible solution is optimal.

5. Select a q such that rq < 0

6. If no yi'q > 0, stop. -- the problem is unbounded; else, calculate $ p=argmin_{i}\left \{ y_{i0}/y_{iq}:y_{iq}>0 \right \} $

7. Update the canonical augmented matrix by pivoting about the (p,q) th element.

8. Go to step 3.


$ \color{blue}\text{Solution 1:} $

   min   x1x2 
   subject to    x1x2 + x3 = 2 
                     x1 + x2 + x4 = 6 

                     $ x_{1},x_{2},x_{3},x_{4}\geq 0 $

$ \begin{matrix} 1 & -1 & 1 & 0 & 2\\ 1 & 1 & 0 & 1 & 6 \\ -1 & -1 & 0 & 0 & 0 \end{matrix} \Rightarrow \begin{matrix} 1 & -1 & 1 & 0 & 2\\ 0 & 2 & -1 & 1 & 4 \\ 0 & -2 & 1 & 0 & 2 \end{matrix} \Rightarrow \begin{matrix} 1 & 0 & \frac{1}{2} & \frac{1}{2} & 4\\ 0 & 1 & -\frac{1}{2} & \frac{1}{2} & 2 \\ 0 & 0 & 0 & 1 & 6 \end{matrix} $

$ \Rightarrow x_{1}=4, x_{2}=2, \text{the maximum value } x_{1}+x_{2}=6 $


$ \color{blue}\text{Solution 2:} $

Get standard form for simplex method   min   x1x2

                                                           subject to    x1x2 + x3 = 2

                                                                             x1 + x2 + x4 = 6

                                                                             $ x_{i}\geq0, i=1,2,3,4 $

$ \begin{matrix} & a_{1} & a_{2} & a_{3} & a_{4} & b\\ & 1 & -1 & 1 & 0 & 2\\ & 1 & 1 & 0 & 1 & 6 \\ c^{T} & -1 & -1 & 0 & 0 & 0 \end{matrix} $      $ \Rightarrow \begin{matrix} 1 & -1 & 1 & 0 & 2\\ 1 & 1 & 0 & 1 & 6 \\ 0 & 0 & 0 & 1 & 6 \end{matrix} \Rightarrow \begin{matrix} 1 & -1 & 1 & 0 & 2\\ 0 & 2 & -1 & 1 & 4 \\ 0 & 0 & 0 & 1 & 6 \end{matrix} \Rightarrow \begin{matrix} 1 & 0 & \frac{1}{2} & \frac{1}{2} & 4\\ 0 & 1 & -\frac{1}{2} & \frac{1}{2} & 2 \\ 0 & 0 & 0 & 1 & 6 \end{matrix} $

$ \therefore \text{the optimal solution to the original problem is } x^{*}= \begin{bmatrix} 4\\ 2 \end{bmatrix} $

The maximum value for   x1 + x2 is 6.


Automatic Control (AC)- Question 3, August 2011

Go to


Back to ECE Qualifying Exams (QE) page

Alumni Liaison

Abstract algebra continues the conceptual developments of linear algebra, on an even grander scale.

Dr. Paul Garrett