13.8.2 Euler-Lagrange equations
The Euler-Lagrange equations for a Lagrangian function
f(x,y,y′) are differential equations which must be satisfied by
extrema of the functional F(y).
The euler_lagrange
command finds the Euler-Lagrange equations for a Lagrangian f.
-
euler_lagrange takes one mandatory argument and
two optional arguments:
-
expr, an expression involving an independent variable, a
dependent variable, and the dependent variable derivative.
- Optionally, indvar, the independent variable (by default x).
- Optionally, depvar, the dependent variable (by default y).
If a function y∈ℝn is required (by default n=1), you can enter
y=(y1,y2,…,yn) as a vector [y1,y2,…,yn].
In that case, y′=(y1′,y2′,…,yn′).
An alternate way to specify the independent and dependent variables
is by replacing both optional arguments by either, for example, y(x) or
[y1(x),y2(x),…,yn(x)].
- euler_lagrange(expr ⟨,indvar,depvar ⟩)
returns the system of differential Euler-Lagrange equations.
For n=1, a single equation is returned:
∂ f/∂ y=d/dx ∂ f/∂ y′
.
In general, n equations are returned:
∂ f/∂ yk=d/dx ∂ f/∂ yk′,
k=1,2,…,n.
The degrees of these differential equations are kept as low as
possible. If, for example, ∂ f/∂ y=0, the
equation ∂ f/∂ y′=K is returned, where
K∈ℝ is an arbitrary constant. Similarly, using the
Hamiltonian
H(x,y,y′)=y′ | | f(x,y,y′)−f(x,y,y′)
|
the Euler-Lagrange equation is simplified in the case n=1 and
∂ f/∂ t=0 to:
since it can be shown that
d/dx H(y,y′,x)=0. Therefore the
Euler-Lagrange equations, which are generally of order two in y, are
returned in a simpler form of order one in the aforementioned cases. If
n=1 and ∂ f/∂ x=0, then both
equations are returned, each of them being sufficient to determine y (one of the
returned equations is usually simpler than the other).
Examples
Minimize the functional F for 0<a<b and
f(x,y,y′)=x2 y′(x)2+y(x)2.
eq:=euler_lagrange(x^2*diff(y(x),x)^2+y^2) |
This can be solved by assuming y(x)=xr for some r∈ℝ.
solve(subs(eq,y(x)=x^r),r) |
The same pair of solutions is also returned by
the kovacicsols command (see Section 13.4.3):
assume(x>=0):;
kovacicsols(y''=(y-2x*y')/x^2,x,y) |
You can conclude that y=C1 x−√5+1/2+C2 x√5+1/2.
The values of C1 and C2 are determined from the boundary conditions.
Finally, to prove that f is convex:
convex(x^2*diff(y(x),x)^2+y^2,y(x)) |
Therefore, y minimizes F on [a,b].
Find the function y in
{y∈ C1[1/2,1]:y(1/2)=−√3/2, y(1)=0}
which minimizes the functional
To obtain the corresponding Euler-Lagrange equation:
eq:=euler_lagrange(sqrt(1+diff(y(x),x)^2)/x) |
The sought solution is the function of the above form which satisfies
the boundary conditions.
y0,c:=sol[0],[K_3,c_0]:;
v:=solve([subs(y0,x=1/2)=-sqrt(3)/2,subs(y0,x=1)=0],c) |
y0:=normal(subs(y0,c,v[0]) |
If the integrand in F(y) is convex, then y0(x)=−√1−x2
is a minimizer for F. Indeed:
convex(sqrt(1+y'^2)/x,y(x)) |
You can similarly find the minimizer for
F(y)= | ∫ | | ⎛
⎝ | 2 sin(x) y(x)+y′(x)2 | ⎞
⎠ | dx
|
where y∈ C1[0,π] and y(0)=y(π)=0.
eq:=euler_lagrange(2*sin(x)*y(x)+diff(y(x),x)^2) |
dsolve(eq and y(0)=0 and y(pi)=0,x,y) |
The above function is the sought minimizer as the integrand
2 sin(x) y(x)+y′(x)2
is convex:
convex(2*sin(x)*y(x)+diff(y(x),x)^2,y(x)) |
Minimize the functional F(y)=∫01(y′(x)4−4 y(x)) dx
on C1[0,1] with boundary conditions y(0)=1 and y(1)=2.
First, solve the associated Euler-Lagrange equation:
eq:=euler_lagrange(y'^4-4y,x,y) |
|
| ⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣ | 3 | ⎛
⎜
⎜
⎝ | | y | ⎛
⎝ | x | ⎞
⎠ | ⎞
⎟
⎟
⎠ | | +4 y | ⎛
⎝ | x | ⎞
⎠ | =K6, | | y | ⎛
⎝ | x | ⎞
⎠ | =− | | ⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦ |
| | | | | | | | | | |
|
dsolve(eq[1] and y(0)=1 and y(1)=2,x,y) |
|
| ⎡
⎢
⎢
⎣ | − | | | ⎛
⎝ | −x+1.52832425067 | ⎞
⎠ | | +2.32032831141 | ⎤
⎥
⎥
⎦ |
| | | | | | | | | | |
|
Next, check if the integrand in F(y) is convex:
Hence the minimizer is
y0(x)=−3/4 (1.52832425067−x)4/3+2.32032831141, 0≤ x≤ 1.
Find Euler-Lagrange equations for a bivariate functional F:
euler_lagrange(sqrt(x'(t)^2+y'(t)^2),[x(t),y(t)]) |
|
| ⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣ |
| |
|
| |
⎛
⎜
⎜
⎝ | | x | ⎛
⎝ | t | ⎞
⎠ | ⎞
⎟
⎟
⎠ | | + | ⎛
⎜
⎜
⎝ | | y | ⎛
⎝ | t | ⎞
⎠ | ⎞
⎟
⎟
⎠ | |
|
|
|
| =K0,
| |
|
| |
⎛
⎜
⎜
⎝ | | x | ⎛
⎝ | t | ⎞
⎠ | ⎞
⎟
⎟
⎠ | | + | ⎛
⎜
⎜
⎝ | | y | ⎛
⎝ | t | ⎞
⎠ | ⎞
⎟
⎟
⎠ | |
|
|
|
| =K1
| ⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦ |
| | | | | | | | | | |
|
where K0,K1∈ℝ are arbitrary constants
(note that these symbols are generated automatically).
It can be proven that if f is convex (as a function of three
independent variables, see Section 13.7.8), then a solution y
to the Euler-Lagrange equations minimizes the functional F.