D-admissible directions and extermal points

  • Thread starter Thread starter braindead101
  • Start date Start date
  • Tags Tags
    Points
Click For Summary
SUMMARY

The discussion focuses on the optimization problem involving the functional J(y) = ∫(a to b) sin^3(x) + y(x)^2 dx, constrained by the condition G(y) = ∫(a to b) y(x) dx = 1. The user seeks to identify D-admissible directions for J and to find all possible local external points for J on the set D. The calculations presented utilize the limit definition of the derivative and the Gateaux derivative, leading to the expression for D-admissible directions as y + Ev ∈ D, where E is a small perturbation.

PREREQUISITES
  • Understanding of functional analysis and calculus of variations
  • Familiarity with the concept of Gateaux derivatives
  • Knowledge of integration techniques, specifically in the context of functionals
  • Experience with optimization problems under constraints
NEXT STEPS
  • Study the properties of D-admissible directions in optimization problems
  • Learn about the Gateaux derivative and its applications in functional analysis
  • Explore the method of Lagrange multipliers for constrained optimization
  • Investigate examples of local external points in variational problems
USEFUL FOR

Mathematicians, optimization specialists, and students in advanced calculus or functional analysis seeking to deepen their understanding of variational methods and constrained optimization techniques.

braindead101
Messages
158
Reaction score
0
Let X = C[a,b], J(y) = integ(a to b) sin^3(x) + y(x)^2 dx and D={yEX; integ(a to b) y(x)dx = 1}
(a) what are the D-admissible directions for J?
(b) Find all possible (local) external points for J on D?

so far i have:
(let e be epsilon)
lim e->0 J(y+ev) - J(y) / e
= lim e->0 integ (a to b) [ sin^3(x) + (y+ev)^2(x)dx - J(y) ] / e
= lim e->0 integ (a to b) [2ey(x)v(x) + e^2v(x)^2) dx ] / e
= 2 integ(a to b) y(x)v(x) dx
A similar example was done in class, so i just copied it for this question.
G(y) = integ(a to b) y(x) dx = 1
gateaux G(y;v) = integ (a to b) y(x)v(x) dx

is this right so far? how do i go about answering (a) and (b).
 
Physics news on Phys.org
i was wondering if the D-admissible directions for J is just
y+Ev E D
so,
integ (a-b) (y+Ev)(x) dx
= 1+ E integ(a-b) v(x) dx

can i simplify any further?
 

Similar threads

Replies
5
Views
2K
Replies
4
Views
2K
  • · Replies 18 ·
Replies
18
Views
3K
Replies
2
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 10 ·
Replies
10
Views
2K
Replies
2
Views
1K
  • · Replies 12 ·
Replies
12
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K
  • · Replies 10 ·
Replies
10
Views
3K