Penalty And Augmented Lagrangian Methods To Solve Nonlinear Optimization Prob- Lems

Mathematics Project Topics

Get the Complete Project Materials Now! ยป

Penalty and augmented Lagrangian methods are procedures for approximating constrainedrnoptimization problems by unconstrained optimization problems to and the approximate so-rnlution of the given constrained problem. The idea of replacing a constrained optimizationrnproblem by a sequence of unconstrained problems parametrized by a scalar parameter hasrnplayed a fundamental role in the formulation of algorithms. Even though, penalty meth-rnods are very natural and general; unfortunately, they suffer from a serious drawback: tornapproximate well the solution to constrained problem, we have to work with large penaltyrnparameters, and this inevitably makes the problem of unconstrained minimization of the pe-rnnalized objective very ill-conditioned. Their slow rates of convergence due to ill-conditioningrnof the associated Hessian led researchers to pursue other approaches, augmented Lagrangianrnmethods. The main advantage of the augmented Lagrangian methods is that they allowrnto approximate well the solution to a constrained problem by solutions of unconstrainedrn(and penalized) auxiliary problems without pushing the penalty parameter to inifnity; as arnresult, the auxiliary problems remain reasonably conditioned even when we are seeking forrnhigh-accuracy solutions. In this paper we will see how all this works.rnKeywords: Unconstrained optimization, constrained optimization, penalty and barrierrnmethods, augmented Lagrangian methods.

Get Full Work

Report copyright infringement or plagiarism

Be the First to Share On Social



1GB data
1GB data

RELATED TOPICS

1GB data
1GB data
Penalty And Augmented Lagrangian Methods To Solve Nonlinear Optimization Prob- Lems

150