Some experiments on heuristic code selection versus numerical performance in nonlinear programming

Some situations are described in this paper which require the exploitation of heuristics either to solve a nonlinear programming problem or to analyse results. A possibility to implement heuristc knowledge for selecting a suitable algorithm depending on available problems data and information retrieved from the user, is introduced. A rule-based module based on uncertainties was implemented within the interactive system EMP. Case studies are presented which outline on the one hand the heuristic recommendation of an optimization code by EMP and the numerical results achieved by some algorithms on the other hand.