EE4047_Part1

# EE4047_Part1 - x f sin 2 1 ) ( 2 − = x x g x x x f x g x...

This preview shows pages 1–5. Sign up to view the full content.

City University of Hong Kong Part 1: Optimization - Problems and Classical Methods Optimization Problems ± Find the “best” solution from all feasible solutions for a problem ± Best for a function, cost, … ± Examples: ± Dimensions of Antenna ± Wireless network layout ± Parameters of controllers and systems ± Facial reconstruction of a criminal ±

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Condition of minimization ± Function with single variable x 0 x m x c f(x) x X * Local and Global minimum ± A function has a local minimum at the value x 0 if there exists a positive value δ such that if ± A function f(x) has a global minimum at x * if for all values of x ± f’(x) = 0 for both cases; ± f’’(x)> 0 for minimum points ( ) ( ) 0 0 , x f x f x x < δ () () * x f x f N.B.: f’(x) represents the differentiation of the function f w.r.t. x
Example x x x f sin 2 1 ) ( 2 = Newton’s method O x 1 T x 0 P θ tan θ = g’(x 0 ) TA=PA/ tan θ TA = g(x 0 )/g’(x 0 ) x 1 = x 0 -g(x 0 )/g’(x 0 ) x 2 = x 1 -g(x 1 )/g’(x 1 ) : : X r+1 = x r -g(x r )/g’(x r ) x y=g(x) N.B.: g(x) = f’(x) represents the gradient of the function f x 2 A

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Example ± Find the minimum of ± We have x x
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: x f sin 2 1 ) ( 2 − = x x g x x x f x g x x x f sin 1 ) ( cos ) ( ) ( sin 2 1 ) ( 2 + = ′ − = ′ = − = 0.000000006 0.00009461 0.02710331-0.37758256 g(x r ) 1.673612032 1.673653810 1.685450631 1.479425538 g’(x r ) 0.739085133 0.739085134 4 0.739085134 0.739141666 3 0.739141666 0.755222417 2 0.755222417 0.5 1 x r+1 x r Comments ± Fast and simple ± may not be global ± Information about gradient is needed (differentiable) ± Can be extended to n-variable problem Update rule: where λ i is the value of λ that minimize and ( ) r i r r f x x x ∇ − = + λ 1 ( ) ( ) ( ) r r f f x x ∇ − = φ ( )         ∂ ∂ ∂ ∂ ∂ ∂ = ∇ n x f x f x f f , , , 2 1 L x Random Search ± Random generate sufficient number of candidates and find the optimal one ± Comments: ± A large number of candidates needed ± Commonly miss the best one...
View Full Document

## This note was uploaded on 01/11/2011 for the course EE 4047 taught by Professor Kitsangtsang during the Fall '09 term at City University of Hong Kong.

### Page1 / 5

EE4047_Part1 - x f sin 2 1 ) ( 2 − = x x g x x x f x g x...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online