Optimisation Ada software ?

[Jean Claude Berges <Jean-Claude.Berges@CST.CNES.FR>]

I am looking for an Ada procedure which
realizes the minimisation of the quadratic criterion
C= Q(S) by optimising the elements of E.

E and S are vectors of floats, and are linked by
the relation S=function_F(E);

Fonction_F is coded in Ada. It is non-linear, not continuous,
but quite regular. Whe hope there is no local minima. Speed of
variations might be very different on the different axes.

Speed of convergence must be high, precision is not critical.

Please make a copy of your answer to serge@lune.cst.cnes.fr

Thanks a lot.

Serge Potteck