Hi,
by accident, I made a copy and paste mistake in the estimated_params section. I had included one parameter, a standard deviation of a shock, twice. Dynare then estimated the model without any error message and gave me two different values for the estimated standard deviation at the posterior mode. I have no idea what Dynare did internally, but obviously having two values simultaneously makes no sense.
Hence, I would suggest to add an additional check in the preprocessing if the parameters to be estimated are uniquely specified.
Best regards,
Johannes