Calculus (3rd Edition)

Published by W. H. Freeman
ISBN 10: 1464125260
ISBN 13: 978-1-46412-526-3

Chapter 15 - Differentiation in Several Variables - 15.8 Lagrange Multipliers: Optimizing with a Constraint - Exercises - Page 831: 16

Answer

(a) Using Theorem 1, we show that there is a unique point $P = \left( {1,1} \right)$ on $g\left( {x,y} \right) = 1$ where $\nabla {f_P} = \lambda \nabla {g_P}$ for $\lambda = 2$. (b) $f$ has a local maximum at $P$ subject to the constraint. (c) Figure 14 suggests that $f\left( P \right)$ is a global maximum subject to the constraint, within the range as is shown in this figure.

Work Step by Step

(a) We have $f\left( {x,y} \right) = {x^3} + xy + {y^3}$ and the constraint $g\left( {x,y} \right) = {x^3} - xy + {y^3} = 1$. Using Theorem 1, the Lagrange condition $\nabla f = \lambda \nabla g$ yields $\left( {3{x^2} + y,3{y^2} + x} \right) = \lambda \left( {3{x^2} - y,3{y^2} - x} \right)$ So, (1) ${\ \ \ \ }$ $3{x^2} + y = \lambda \left( {3{x^2} - y} \right)$, ${\ \ }$ $3{y^2} + x = \lambda \left( {3{y^2} - x} \right)$ Since $\left( {0,0} \right)$ does not satisfy the constraint $g\left( {x,y} \right)=1$, we may have the following cases: Case 1: $x=0$, $y \ne 0$ Equation (1) becomes $y = - \lambda y$, ${\ \ \ }$ $3{y^2} = 3\lambda {y^2}$ The first equation gives $\lambda = - 1$. But $\lambda = - 1$ does not satisfy the second equation. Therefore, we conclude that there is no solution for this case. Case 2: $x \ne 0$, $y=0$ Equation (1) becomes $3{x^2} = 3\lambda {x^2}$, ${\ \ \ }$ $x = - \lambda x$ The first equation gives $\lambda = 1$. But $\lambda = 1$ does not satisfy the second equation. Therefore, we conclude that there is no solution for this case. Case 3: $x \ne 0$, $y \ne 0$ From equation (1) we obtain $\lambda = \frac{{3{x^2} + y}}{{3{x^2} - y}} = \frac{{3{y^2} + x}}{{3{y^2} - x}}$. $\left( {3{x^2} + y} \right)\left( {3{y^2} - x} \right) = \left( {3{x^2} - y} \right)\left( {3{y^2} + x} \right)$ $9{x^2}{y^2} + 3{y^3} - 3{x^3} - xy = 9{x^2}{y^2} - 3{y^3} + 3{x^3} - xy$ $6{y^3} = 6{x^3}$ $y = x$ Substituting $y = x$ in the constraint $g\left( {x,y} \right) = {x^3} - xy + {y^3} = 1$ gives ${x^3} - {x^2} + {x^3} = 1$ $2{x^3} - {x^2} - 1 = 0$ $\left( {x - 1} \right)\left( {2{x^2} + x + 1} \right) = 0$ $x - 1 = 0$ ${\ \ }$ or ${\ \ }$ $2{x^2} + x + 1 = 0$ However, the solutions for $2{x^2} + x + 1 = 0$ is $x = \frac{{ - 1 \pm \sqrt {{1^2} - 4\cdot2\cdot1} }}{4}$, which are complex numbers. Therefore, for real solution, there is only one solution at $x=1$. Using $y=x$, we obtain $y=1$. Using $\lambda = \frac{{3{x^2} + y}}{{3{x^2} - y}}$, we obtain $\lambda = 2$. Thus, the critical point is $P = \left( {1,1} \right)$. Hence, there is a unique point $P = \left( {1,1} \right)$ on $g\left( {x,y} \right) = 1$ where $\nabla {f_P} = \lambda \nabla {g_P}$ for $\lambda = 2$. (b) Suppose we move toward $P$ along the constraint curve $g\left( {x,y} \right) = 1$ (the red curve in Figure 14), $f$ is increasing (either we approach $P$ from the left or from the right) until we arrive at $P$, where $\nabla {f_P}$ is orthogonal to the constraint curve $g\left( {x,y} \right) = 1$. Once at $P$, we cannot increase $f$ further without leaving the constraint curve. Thus, $f$ has a local maximum at $P$ subject to the constraint. (c) In Figure 14 we see that there is no other point $\left( {x,y} \right)$ along the constraint curve $g\left( {x,y} \right) = 1$, where the value $f\left( {x,y} \right)$ is larger than the value of $f$ at $P$. Therefore, Figure 14 suggests that $f\left( P \right)$ is a global maximum subject to the constraint, within the range as is shown in this figure.
Update this answer!

You can help us out by revising, improving and updating this answer.

Update this answer

After you claim an answer you’ll have 24 hours to send in a draft. An editor will review the submission and either publish your submission or provide feedback.