I use Abaqus2Matlab to determine material parameters with an inverse analysis.
I use the following optimization procedure:
- Initialization of the variable x = [Eta1, E2, Eta2]
- Creation of the subroutine with the variable x used in the Abaqus job
- Abaqus job run
- Extraction of the node displacements needed and the time increment
- Comparison with the experimental data.
Every parts of the code works, but the value of the parameter x is not changing during the optimization procedure.
Do you have an explication to this problem? Thanks.
I am attaching my Matlab files:
Dear marthe.loiseau,
Please check if the result of the Abaqus analysis changes while you change one of the three design variables of the optimization problem. The values of the objective function (Cost) must vary even for infinitesimal variations of the design variables in order for fmincon to work properly.
Let me know about the above.
If you wish, you can send me your files through p.m. at gpapazafeiropoulos@yahoo.gr so that I can check more effectively what is going on.