0% found this document useful (0 votes)
12 views4 pages

ADE Exam Spring 2023 Overview

The document outlines the ADE Exam for Spring 2023 from the Department of Mathematics at UCLA, consisting of various mathematical problems related to dynamical systems, Legendre equations, energy functionals, nonlinear partial differential equations, and the Korteweg-de Vries equation. Each problem requires analysis, derivation of equations, and proofs regarding stability, uniqueness, and properties of solutions. The exam tests knowledge in advanced mathematical concepts and techniques relevant to the topics presented.

Uploaded by

3089688246
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views4 pages

ADE Exam Spring 2023 Overview

The document outlines the ADE Exam for Spring 2023 from the Department of Mathematics at UCLA, consisting of various mathematical problems related to dynamical systems, Legendre equations, energy functionals, nonlinear partial differential equations, and the Korteweg-de Vries equation. Each problem requires analysis, derivation of equations, and proofs regarding stability, uniqueness, and properties of solutions. The exam tests knowledge in advanced mathematical concepts and techniques relevant to the topics presented.

Uploaded by

3089688246
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

ADE Exam, Spring 2023

Department of Mathematics, UCLA

1. [10 points]

(a) Consider the dynamical system

dx
= y,
dt
dy
= −x − ϵx2 y , (x, y) ∈ R2 , (1)
dt
where ϵ ≥ 0 is a parameter.
Determine the stability of the equilibrium point at (0, 0).
(b) Consider the dynamical system

dx
= y,
dt
dy
= x − x3 − δy + x2 y , (x, y) ∈ R2 , (2)
dt
where δ > 0 is a constant.
Determine the equilibrium points of (2), and use linear stability analysis to classify
√ their
type and (when possible) their stability. Show that the two vertical lines x = ± δ divide
the phase plane into three regions such that a periodic orbit cannot exist entirely in one
of these regions.

2. [10 points] Consider the Legendre equation

d2 y dy
(1 − x2 ) − 2x + ℓ(ℓ + 1)y = 0 , −1 ≤ x ≤ 1 , (3)
dx2 dx
where ℓ ≥ 0 is an integer. Let Pℓ denote the solution of (3) that satisfies Pℓ (1) = 1.

(a) Show that x = 1 is a regular singular point. Find the indicial equation and indicial
exponents, and find the leading terms of the series expansion at x = 1 for two linearly
independent solutions. Use them to explain why the condition Pℓ (1) = 1 is sufficient to
uniquely determine Pℓ .
P∞
(b) Derive a recursion relation for the coefficients of the series expansion y(x) = k=0 ak xk
for solutions of (3). Using this relation, show that Pℓ is a polynomial that (i) consists
only of even powers when ℓ is even and (ii) consists only of odd powers when ℓ is odd.
2

(c) Using the Rodrigues formula

1 dℓ h 2 ℓ i
Pℓ (x) = x − 1 , (4)
2ℓ ℓ! dxℓ
or otherwise, show that Pℓ satisfies the orthogonality relation
Z 1
Pℓ (x)Pm (x) dx = 0 , (5)
−1

and determine the value of the integral (5) when ℓ = m.

3. [10 points] Consider the energy functional E[u], which is defined for u ∈ C 2 (D) by
Z
1
|∇u|2 + u2 dn x ,

E[u] =
2 D
where D ⊂ Rn is a bounded and open set. Assume that u = g(x) is known on ∂D.
(a) Derive the partial differential equation that is satisfied by the minimizer of E. Starting
from the minimization principle, prove that solutions of this PDE are unique.
(b) Suppose that n = 1, D = (−1, 1), and u(−1) = u(1) = 1. Find an approximate solution
of your PDE from (a) that takes the form u = 1 + A(1 − x2 ). That is, find the value of
A that minimizes the energy functional.

4. [10 points] Consider the nonlinear partial differential equation

ut = ∆u − u3 , x ∈ D , 0 < t < T ,

where D ⊂ Rn is a bounded and open set. You may assume that solutions exist and are
C 2,1 (D × (0, T )) ∩ C(D̄ × [0, T ]). Show that the solutions of the PDE are unique.

5. [10 points] Consider the one-dimensional partial differential equation


1
ut + u2x = 0 , − ∞ < x < ∞,
2
with initial condition

u(x, 0) = 0 , x < 0; u(x, 0) = 1 , x>0

and boundary conditions

u → 0 as x → −∞ and u → 1 as x → ∞ .
3

(a) Show that the PDE does not have a traveling-wave solution that is compatible with
these boundary conditions, even when the derivatives are interpreted in the sense of
distributions.
(b) Derive the weak solution of the PDE.
[Note: You do not need to derive the Hopf–Lax formula, but you should state the formula
carefully if you use it.]

6. [10 points] Consider the partial differential equation

ut + uux = −u ,

with initial condition

u(x, 0) = 1 , x < 0 ; u(x, 0) = 0 , x > 0 .

(a) Show that for a smooth solution, the PDE can be written in terms of the “characteristic”
variable x = ξ(t) as
d dξ
u(ξ(t)) = −u(ξ(t)) ; = u.
dt dt
(b) Suppose that u(0, 0) takes values α ∈ (0, 1). Solve for the characteristics ξα (t) starting
from x = 0 with the initial value u = α.
(c) Using the result from (b), solve the Riemann problem with the initial condition above.
Write your answer in terms of the Eulerian variables x and t.

7. [10 points] Consider the Korteweg–de Vries (KdV) equation

ut + uxxx + 6uux = 0 , x ∈ R, t > 0, u(x, 0) = f (x) .

Assume that the function u(x, t) and all of its derivatives vanish as |x| → ∞.
(a) Show that the following are conserved quantities in time:
Z ∞ Z ∞ 
2 1 2 3
u (x, t) dx ; ux (x, t) − u (x, t]) dx .
−∞ −∞ 2

(b) Show that the KdV equation does not preserve positivity of the solution by constructing
an initial condition f that is nonnegative for which the solution becomes negative at a
later time.
[Hint: Consider a local minimum for which the third derivative in space is nonzero.]
4

8. [10 points] Solve the initial-value problem

utt − 2uxt − 15uxx = 0 ,

with u(x, 0) = g(x) and ut (x, 0) = h(x).

[Hint: Consider factoring the differential operator.]

Common questions

Powered by AI

Conservation properties of the KdV equation, such as the conservation of the integral of u^2(x, t) and of the integral of 1/2 u_x^2 - u^3, are crucial because they imply the stability of the soliton solutions over time. These conserved quantities indicate that certain behaviors (like wave amplitude) persist unchanged, which is essential for analyzing stability and long-term behavior. However, despite these properties, non-positivity can be demonstrated by constructing initial conditions where the third spatial derivative is non-zero, leading to changes in sign due to the wave's evolution, such as at a local minimum where positivity is not preserved .

The orthogonality relation for Legendre polynomials states that for different integers ℓ and m, the integral from -1 to 1 of the product Pℓ(x) and Pm(x) is zero. This orthogonality property extends the concept of polynomials being orthogonal over a weighted interval, akin to Gram-Schmidt orthogonality in vector spaces. When ℓ equals m, the value of the integral ∫ from -1 to 1 of Pℓ(x)^2 dx is given by 2/(2ℓ + 1).

The significance of a regular singular point for a differential equation like the Legendre equation is that it allows the solution to be expressed as a Frobenius series, which converges near that point. At a regular singular point, the equation may have solutions that include logarithmic terms. For the Legendre equation, the indicial equation is determined by substituting a power series solution y(x) = (x-1)^r Σa_n (x-1)^n into the differential equation and ensuring that the lowest power of (x-1) vanishes. This results in the indicial equation r(r-1) = 0, giving indicial exponents r = 0 and r = 1 .

The uniqueness of the solutions to the PDE is affirmed through the method of characteristics, which transforms the PDE into ordinary differential equations along curves (or characteristics) in the space-time domain. For the given equation, the characteristic equations are derived as d(ξ(t))/dt = u and d(u(ξ(t)))/dt = -u, which describe how u and the space variable x evolve over time. These yield a system where solutions for u at any point (x, t) are determined by following these characteristic lines, ensuring uniqueness when initial conditions are appropriately set .

Choosing a trial function u = 1 + A(1-x^2) is strategic for minimizing the energy functional E[u] because it maintains the boundary conditions while being amenable to calculus operations, particularly differentiation and integration. This form allows for easy computation of derivatives needed for calculating the gradient and Laplacian involved in E, simplifying the evaluation of the functional. The parameter A encapsulates the degree of deviation from a trivial constant solution, and by calculating E as a function of A, a minimizer A* can be determined that yields the minimum energy, fulfilling the variational principle .

The analytical approach involves factoring the differential operator in the PDE. This can be done by finding functions that simplify the expression, possibly by a change of variables or by transforming the operator into a product of first-order operators. In this case, one aims to separate the variables t and x, often by applying techniques such as Fourier transforms, method of characteristics, or by simplifying it to a wave equation form .

The Rodrigues formula simplifies the proof of orthogonality of Legendre polynomials by providing a straightforward, explicit representation of these polynomials that is differentiable, easily manipulated, and adjustable through the order ℓ. The formula implies an inherent weight function of (1-x^2)^ℓ, which naturally vanishes at the endpoints x=±1, ensuring the polynomials are orthogonal over the interval [-1,1]. This orthogonality with respect to the standard weight function on this interval implies that the polynomials form a complete orthonormal basis in the space of square-integrable functions on [-1, 1], facilitating their application in approximating functions .

A weak solution can be derived using integral methods or distributional derivatives to accommodate discontinuities or non-smooth behavior in u(x,t). The traveling wave solution fails because the boundary conditions imply a discontinuity between u=0 and u=1, which cannot be reconciled with a steady profile moving with constant speed. Thus, instead of smooth traveling waves, the solution is characterized by shock waves or discontinuities, where characteristics intersect or break down, specifically handled by defining a weak solution that respects the PDE in an integral sense across these discontinuities .

The non-existence of periodic orbits within the specific phase plane regions, defined by the lines x = ±√δ, suggests complex and potentially chaotic behavior in the dynamics of the system as opposed to stable limit cycles. This segmentation indicates that trajectories may cross these boundaries, leading to a different qualitative behavior outside each defined region. The absence of bounded periodic orbits in any single region means that solutions will not repeat exactly, which could imply sensitivity to initial conditions characteristic of more chaotic systems or the possibility of irregular motion converging to, or spiraling away from, equilibrium points .

The parameter δ influences the stability and type of the equilibrium points by affecting the linear stability analysis of the system. By setting dx/dt = 0 and dy/dt = 0 and solving for x and y, one finds the equilibrium points, which are influenced by the critical number δ because it modifies the damping term in the equation. The nature (stability) of these equilibrium points can be assessed by examining the Jacobian matrix's eigenvalues evaluated at those points. Changes in δ alter these eigenvalues, thus modifying the stability and bifurcation nature of the equilibrium points .

You might also like