On The Penalty Method In Optimization Theory And Numerical Realization.

In this thesis, we deal with a theoretical basis of optimization in general and on numericalrnmethods in particular. Here the penalty methods are in the foreground . Some generalizationsrnfor theorems into arbitrary metric space with mild assumptions (Theorem 2.1.3 and 3.1.3) arerndone. In addition, a numerical computation is presented and tested, with the code written inrnMathematica 4.1 . Comparisons are made for different choice of the penalty function withrnrespect to convergence speed and objective value.

Get Full Work

Report copyright infringement or plagiarism