CONVEX NON-LINEAR PROGRAMMING
Methods
Approach sequential
consists
Substitution
using
Aproximations
solving
Linear Methods
Frank and Wolfe
consist in
Find Solution Feasible
through
Approach Linear
Gradient Method
solves
Steps
Problems without Restrictions
- x=x(0), k=0
- Choose dk
- Do search linear ak
- x(k+1) = x + ak * dk
- Test of convergence x(k+1)
Converge
No Converge
Ends
return 2
Consist in
Function
Objective
must be
with
Restrictions
Characterized by
Own or Not
Restrictions
it has
is unrestricted
Must be Convex
Must be Linear
Non - Linear
where
as
Convex
Minimize
Concave
Maximize
Quadratic
Separable
Aplications
Design of
Transportation Networks
Production and
Addressing Network
Production in
Factories
Software
MOSEK
Mathematical Programming
Language (MPL)
