Conditions for Integral Functionals to have same Euler's Equations
Jump to navigation
Jump to search
This article needs to be linked to other articles. You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by adding these links. To discuss this page in more detail, feel free to use the talk page. When this work has been completed, you may remove this instance of {{MissingLinks}} from the code. |
Theorem
Let $\mathbf y$ be a real $n$-dimensional vector-valued function.
Let $\map F {x, \mathbf y, \mathbf y'}$, $\map \Phi {x, \mathbf y}$ be real functions.
Let $\Phi$ be twice differentiable.
Let:
\(\ds \Psi\) | \(=\) | \(\ds \frac {\d \Phi} {\d x}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \frac {\partial \Phi} {\partial x} + \sum_{i \mathop = 1}^n \frac {\partial \Phi} {\partial y_i} y_i'\) |
Let $J_1$, $J_2$ be functionals such that:
- $\ds J_1 \sqbrk {\mathbf y} = \int_a^b \map F {x, \mathbf y, \mathbf y'} \rd x$
- $\ds J_2 \sqbrk {\mathbf y} = \int_a^b \paren {\map F {x, \mathbf y, \mathbf y'} + \map \Psi {x, \mathbf y, \mathbf y'} } \rd x$
Then $J_1$ and $J_2$ have same Euler's Equations.
Proof
According to Necessary Condition for Integral Functional to have Extremum for given function/Dependent on N Functions:
Euler's Equations for functional $J_1$ are:
- $\ds F_{\mathbf y} - \frac \d {\d x} F_{\mathbf y'} = 0$
Equivalently, for $J_2$ we have
\(\ds \paren {F_{\mathbf y} + \Psi_{\mathbf y} } - \map {\frac \d {\d x} } {F_{\mathbf y'} + \Psi_{\mathbf y'} }\) | \(=\) | \(\ds \paren {F_{\mathbf y} - \frac \d {\d x} F_{\mathbf y'} } + \paren {\Psi_{\mathbf y} - \frac \d {\d x} \Psi_{\mathbf y'} }\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 0\) | condition for the existence of extremum |
Furthermore:
\(\ds \Psi_{\mathbf y}\) | \(=\) | \(\ds \frac {\partial^2 \Phi} {\partial \mathbf y \partial x} + \frac \partial {\partial \mathbf y} \sum_{j \mathop = 1}^n \frac {\partial \Phi} {\partial y_j} y_j'\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \frac {\partial^2 \Phi} {\partial \mathbf y \partial x} + \sum_{j \mathop = 1}^n \frac {\partial^2 \Phi} {\partial \mathbf y \partial y_j} y_j'\) |
\(\ds \Psi_{\mathbf y'}\) | \(=\) | \(\ds \frac {\partial^2 \Phi} {\partial \mathbf y' \partial x} + \frac \partial {\partial \mathbf y'} \sum_{j \mathop = 1}^n \frac {\partial \Phi} {\partial y_j} y_j'\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 0 + \sum_{j \mathop = 1}^n \paren {\frac {\partial^2 \Phi} {\partial \mathbf y' \partial y_j} y_j' + \frac {\partial \Phi} {\partial y_j} \frac {\partial y_j'} {\partial \mathbf y'} }\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \frac {\partial \Phi} {\partial \mathbf y}\) | as $\frac {\partial y_i'} {\partial y_j'} = \delta_{i j} $, where $\delta_{i j}$ is the Kronecker Delta |
\(\ds \frac {\d \Psi_{\mathbf y'} } {\d x}\) | \(=\) | \(\ds \frac {\partial \Psi_{\mathbf y'} } {\partial x} + \sum_{j \mathop = 1}^n \frac {\partial \Psi_{\mathbf y'} } {\partial y_j} \frac {\d y_j} {\d x} + \sum_{j \mathop = 1}^n \frac {\partial \Psi_{\mathbf y'} } {\partial y_j'} \frac {\d y_j'} {\d x}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \frac {\partial^2 \Phi} {\partial x \partial \mathbf y} + \sum_{j \mathop = 1}^n \frac {\partial^2 \Phi} {\partial \mathbf y \partial y_j} y_j'\) |
Since $\Phi$ is twice differentiable, by Schwarz-Clairaut Theorem partial derivatives commute and:
- $\Psi_{\mathbf y} - \dfrac \d {\d x} \Psi_{\mathbf y'} = 0$
Therefore, $J_1$ and $J_2$ have same Euler's Equations.
$\blacksquare$
Sources
- 1963: I.M. Gelfand and S.V. Fomin: Calculus of Variations ... (previous) ... (next): $\S 2.9$: The Fixed End Point Problem for n Unknown Functions