OCR D1 2013 June — Question 6 21 marks

Exam BoardOCR
ModuleD1 (Decision Mathematics 1)
Year2013
SessionJune
Marks21
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicLinear Programming
TypeInteger solution optimization
DifficultyModerate -0.3 This is a standard D1 linear programming question covering routine techniques: graphical representation, vertex evaluation, integer programming by enumeration, and basic Simplex algorithm. While multi-part and requiring careful execution, all methods are textbook procedures with no novel problem-solving required. The integer programming part (iii) involves systematic checking rather than insight. Slightly easier than average A-level due to being a methodical application of learned algorithms.
Spec7.06a LP formulation: variables, constraints, objective function7.06d Graphical solution: feasible region, two variables7.07a Simplex tableau: initial setup in standard format7.07b Simplex iterations: pivot choice and row operations

6 Consider the following linear programming problem.
Maximise\(P = 5 x + 8 y\),
subject to\(3 x - 2 y \leqslant 12\),
\(3 x + 4 y \leqslant 30\),
\(3 x - 8 y \geqslant - 24\),
\(x \geqslant 0 , y \geqslant 0\).
  1. Represent the constraints graphically. Shade the regions where the inequalities are not satisfied and hence identify the feasible region.
  2. Calculate the coordinates of the vertices of the feasible region, apart from the origin, and calculate the value of \(P\) at each vertex. Hence find the optimal values of \(x\) and \(y\), and state the maximum value of the objective.
  3. Suppose that, additionally, \(x\) and \(y\) must both be integers. Find the maximum feasible value of \(y\) for every feasible integer value of \(x\). Calculate the value of \(P\) at each of these points and hence find the optimal values of \(x\) and \(y\) with this additional restriction. Tom wants to use the Simplex algorithm to solve the original (non-integer) problem. He reformulates it as follows.
    Maximise\(\quad P = 5 x + 8 y\),
    subject to\(3 x - 2 y \leqslant 12\),
    \(3 x + 4 y \leqslant 30\),
    \(- 3 x + 8 y \leqslant 24\),
    \(x \geqslant 0 , y \geqslant 0\).
  4. Explain why Tom needed to change the third constraint.
  5. Represent the problem by an initial Simplex tableau.
  6. Perform one iteration of the Simplex algorithm, choosing your pivot from the \(\boldsymbol { y }\) column. Show how the pivot row was used to calculate the other rows. Write down the values of \(x , y\) and \(P\) that result from this iteration. Perform a second iteration of the Simplex algorithm to achieve the optimum point.

Question 6:
Part (i)
AnswerMarks Guidance
AnswerMark Guidance
Line through \((4, 4.5)\) and \((6, 3)\)M1 \(\pm 0.2\) on each coordinate
Line through \((4, 0)\) and \((6, 3)\)M1 \(\pm 0.2\) on each coordinate
Line through \((0, 3)\) and \((4, 4.5)\)M1 \(\pm 0.2\) on each coordinate
Shading correct, feasible region (FR) need not be labelled (dependent on all three M marks); need not show shading out \(x < 0\), \(y < 0\)A1 Scroll down to check spare copy; visually impaired candidates have tolerance \(\pm 0.5\) throughout
[4]
Part (ii)
AnswerMarks Guidance
AnswerMark Guidance
\((4, 0)\), \((6, 3)\), \((4, 4.5)\) and \((0, 3)\)M1 Any two vertices of their FR (other than origin), as coordinates or equivalent (from their graph to within \(\pm 0.2\)); or two of the correct vertices (new start)
These four coordinates correct, ignore any extra pointsA1
\((6, 3) \to 54\) and \((4, 4.5) \to 56\) (both seen, not implied); or \(P = 5x + 8y\) calculated for two of their vertices with \(x > 0\), \(y > 0\)M1
Optimal when \(x = 4\), \(y = 4.5\), \(P = 56\)A1 \((4, 4.5)\) indicated (e.g. arrowed) and \((4, 4.5) \to 56\) seen (cao)
[4]
Part (iii)
AnswerMarks Guidance
AnswerMark Guidance
Four (\(x\), max \(y\)) pairs, in any form (follow through their FR)M1
All 7 pairs correct, and no extras (cao, not ft)A1
Calculating \(P = 5x + 8y\) correctly for at least three of their listed integer-valued points (not implied from answer \((6, 3)\))M1
Optimal when \(x = 6\), \(y = 3\)A1 \((6, 3)\) written down, in any form (cao)
[4]
Part (iv)
AnswerMarks Guidance
AnswerMark Guidance
So that a slack variable can be added to form an equality, or RHS (value) needs to be non-negativeB1 Allow 'constraints must be \(\leq\)' or 'third constraint was \(\geq\)'; do not allow 'positive' for 'non-negative'
[1]
Part (v)
AnswerMarks Guidance
AnswerMark Guidance
\(4 \times 7\) table with three constraint rows correct (including \(P\) column, although columns need not be labelled)M1 Order of rows and columns may be changed, need not be labelled
Objective row correct; assume blanks mean zeroA1
[2]May credit initial tableau seen in (vi) if no response seen in (v)
Part (vi)
AnswerMarks Guidance
AnswerMark Guidance
Least non-negative ratio in \(y\) column is \(24 \div 8\); pivot on 8 in row 4 of \(y\) columnB1 Correct pivot choice for \(y\) col of their tableau (may be implied)
Row 4 (pr) \(= \text{r4} \div 8\); Row 1 \(= \text{r1} + 8\text{pr}\) or \(\text{r1} + \text{r4}\); Row 2 \(= \text{r2} + 2\text{pr}\), i.e. \(\text{r2} + \frac{1}{4}\text{r4}\); Row 3 \(= \text{r3} - 4\text{pr}\), i.e. \(\text{r3} - \frac{1}{2}\text{r4}\)B1 Correct methods for rows for their positive pivot choice; condone pivot row method not shown; accept reasonable abbreviations, e.g. \(+8\text{pr}\) or \(a+8h\) but not just \(+8\); ft their tableau; condone (r4) for (pr) provided consistent
Augmented tableau with \(P\) value increased and all RHS entries \(\geq 0\); \(x = 0\), \(y = 3\), \(P = 24\)M1, A1 Values of \(x\), \(y\), \(P\) written, ft their tableau for non-negative \(x\), \(y\) and positive \(P\)
Second augmented tableau with \(P\) value increased and all RHS entries \(\geq 0\)M1
Basis columns correspond to \(P = 56\), \(x = 4\), \(y = 4.5\) and a slack variable \(= 9\) (values do not need to be written down); tableau does not need to be correct but does need to correspond to \((4, 4.5)\), 56 and 9 from basis cols with 0's and 1 in themA1
[6]
# Question 6:

## Part (i)

| Answer | Mark | Guidance |
|--------|------|----------|
| Line through $(4, 4.5)$ and $(6, 3)$ | M1 | $\pm 0.2$ on each coordinate |
| Line through $(4, 0)$ and $(6, 3)$ | M1 | $\pm 0.2$ on each coordinate |
| Line through $(0, 3)$ and $(4, 4.5)$ | M1 | $\pm 0.2$ on each coordinate |
| Shading correct, feasible region (FR) need not be labelled (dependent on all three M marks); need not show shading out $x < 0$, $y < 0$ | A1 | Scroll down to check spare copy; visually impaired candidates have tolerance $\pm 0.5$ throughout |
| **[4]** | | |

## Part (ii)

| Answer | Mark | Guidance |
|--------|------|----------|
| $(4, 0)$, $(6, 3)$, $(4, 4.5)$ and $(0, 3)$ | M1 | Any two vertices of their FR (other than origin), as coordinates or equivalent (from their graph to within $\pm 0.2$); or two of the correct vertices (new start) |
| These four coordinates correct, ignore any extra points | A1 | |
| $(6, 3) \to 54$ and $(4, 4.5) \to 56$ (both seen, not implied); or $P = 5x + 8y$ calculated for two of their vertices with $x > 0$, $y > 0$ | M1 | |
| Optimal when $x = 4$, $y = 4.5$, $P = 56$ | A1 | $(4, 4.5)$ indicated (e.g. arrowed) and $(4, 4.5) \to 56$ seen (cao) |
| **[4]** | | |

## Part (iii)

| Answer | Mark | Guidance |
|--------|------|----------|
| Four ($x$, max $y$) pairs, in any form (follow through their FR) | M1 | |
| All 7 pairs correct, and no extras (cao, not ft) | A1 | |
| Calculating $P = 5x + 8y$ correctly for at least three of their listed integer-valued points (not implied from answer $(6, 3)$) | M1 | |
| Optimal when $x = 6$, $y = 3$ | A1 | $(6, 3)$ written down, in any form (cao) |
| **[4]** | | |

## Part (iv)

| Answer | Mark | Guidance |
|--------|------|----------|
| So that a slack variable can be added to form an equality, or RHS (value) needs to be non-negative | B1 | Allow 'constraints must be $\leq$' or 'third constraint was $\geq$'; do not allow 'positive' for 'non-negative' |
| **[1]** | | |

## Part (v)

| Answer | Mark | Guidance |
|--------|------|----------|
| $4 \times 7$ table with three constraint rows correct (including $P$ column, although columns need not be labelled) | M1 | Order of rows and columns may be changed, need not be labelled |
| Objective row correct; assume blanks mean zero | A1 | |
| **[2]** | May credit initial tableau seen in (vi) if no response seen in (v) |

## Part (vi)

| Answer | Mark | Guidance |
|--------|------|----------|
| Least non-negative ratio in $y$ column is $24 \div 8$; pivot on 8 in row 4 of $y$ column | B1 | Correct pivot choice for $y$ col of their tableau (may be implied) |
| Row 4 (pr) $= \text{r4} \div 8$; Row 1 $= \text{r1} + 8\text{pr}$ or $\text{r1} + \text{r4}$; Row 2 $= \text{r2} + 2\text{pr}$, i.e. $\text{r2} + \frac{1}{4}\text{r4}$; Row 3 $= \text{r3} - 4\text{pr}$, i.e. $\text{r3} - \frac{1}{2}\text{r4}$ | B1 | Correct methods for rows for their positive pivot choice; condone pivot row method not shown; accept reasonable abbreviations, e.g. $+8\text{pr}$ or $a+8h$ but not just $+8$; ft their tableau; condone (r4) for (pr) provided consistent |
| Augmented tableau with $P$ value increased and all RHS entries $\geq 0$; $x = 0$, $y = 3$, $P = 24$ | M1, A1 | Values of $x$, $y$, $P$ written, ft their tableau for non-negative $x$, $y$ and positive $P$ |
| Second augmented tableau with $P$ value increased and all RHS entries $\geq 0$ | M1 | |
| Basis columns correspond to $P = 56$, $x = 4$, $y = 4.5$ and a slack variable $= 9$ (values do not need to be written down); tableau does not need to be correct but does need to correspond to $(4, 4.5)$, 56 and 9 from basis cols with 0's and 1 in them | A1 | |
| **[6]** | | |
6 Consider the following linear programming problem.

\begin{center}
\begin{tabular}{ l l }
Maximise & $P = 5 x + 8 y$, \\
 &  \\
subject to & $3 x - 2 y \leqslant 12$, \\
 & $3 x + 4 y \leqslant 30$, \\
 & $3 x - 8 y \geqslant - 24$, \\
 & $x \geqslant 0 , y \geqslant 0$. \\
\end{tabular}
\end{center}

(i) Represent the constraints graphically. Shade the regions where the inequalities are not satisfied and hence identify the feasible region.\\
(ii) Calculate the coordinates of the vertices of the feasible region, apart from the origin, and calculate the value of $P$ at each vertex. Hence find the optimal values of $x$ and $y$, and state the maximum value of the objective.\\
(iii) Suppose that, additionally, $x$ and $y$ must both be integers. Find the maximum feasible value of $y$ for every feasible integer value of $x$. Calculate the value of $P$ at each of these points and hence find the optimal values of $x$ and $y$ with this additional restriction.

Tom wants to use the Simplex algorithm to solve the original (non-integer) problem. He reformulates it as follows.

\begin{center}
\begin{tabular}{ l l }
Maximise & $\quad P = 5 x + 8 y$, \\
subject to & $3 x - 2 y \leqslant 12$, \\
 & $3 x + 4 y \leqslant 30$, \\
 & $- 3 x + 8 y \leqslant 24$, \\
 & $x \geqslant 0 , y \geqslant 0$. \\
\end{tabular}
\end{center}

(iv) Explain why Tom needed to change the third constraint.\\
(v) Represent the problem by an initial Simplex tableau.\\
(vi) Perform one iteration of the Simplex algorithm, choosing your pivot from the $\boldsymbol { y }$ column. Show how the pivot row was used to calculate the other rows. Write down the values of $x , y$ and $P$ that result from this iteration.

Perform a second iteration of the Simplex algorithm to achieve the optimum point.

\hfill \mbox{\textit{OCR D1 2013 Q6 [21]}}