[ADMB Users] Solving a system of non-linear equations inside an optimisation

Mark Payne mpa at aqua.dtu.dk
Thu Apr 28 05:26:23 PDT 2011


I have a problem that seems to be stretching my ADMB ability to breaking point, and was wondering if anyone had some suggestions about how to approach it.

The problem is that I need to solve a system of non-linear equations as part of the exercise of fitting a model to the data. Given an assumed set of parameters, I first need to solve the non-linear system to generate a set of initial conditions. These initial conditions are then integrated in time to give time trajectories, which are compared against the observations to generate an objective function. As you can imagine, each time we change the parameters, we also need to re-solve the non-linear equations - essentially we end up with a nested optimsiation problem

Does anyone have any suggestions what the best way to approach such a problem is?

Best wishes,


More information about the Users mailing list