Is it possible to do a batch run with a Simulink model in rapid accelerator mode and linked to a data dictionary without rebuilding the target every iteration?

Is it possible to do a batch run with a Simulink model in rapid accelerator mode and linked to a data dictionary without rebuilding the target every iteration?

I am changing a variable within the data dictionary for each run. I find that if I set 'RapidAcceleratorUpToDateCheck' to 'off', my model does not seem to update with the changed data dictionary. It does work correctly if I leave 'RapidAcceleratorUpToDateCheck' to 'on.' However, for the latter case that works, when I calculate the checksum each iteration, they appear to be identical. Does this not suggest that I should not have to rebuild?

Answer 1

[Ankit Bhatnagar]

Hi,

Yes, this is expected behavior. Data dictionaries are not tunable parameters and if you select this option to be 'off', it will be faster as there won't be a rebuild but does not satisfy the current workflow.

For applications of this option in parallel simulations, see this:

1. <http://www.mathworks.com/help/simulink/examples/parallel-simulations-using-parsim-parameter-sweep-in-rapid-accelerator-mode.html>