An efficient implementation of optimization algorithms

In implementing any optimization method on a computer, several problems arise: •how to break off the iterative process;•how to determine the satisfactory computed solution and be sure the computer cannot provide a better one;•how to tell whether or not computed solution approximates the mathematical solution, and to what degree of accuracy. New methods of solving these problems are presented in this paper.