I need a help to handle inequality constraints (like x(t)<=1) in Dynare. I was thinking of adding Lagrange multipliers (lamda(t)) and then apply the usual khun-Tucker condition ((1-x(t))*lamda(t)=0) but still I have to make sure that the multiplier lamda(t) is positive for all t. How do I manage this?
What about to penalize utility or profits of agents if the economy breaks the constraint?
Thanks for your answer!
Do you have an idea of how to introduce (automatically) those penalties when writing down the model in the *.mod file for dynare?
Hope that you can help still.
You cannot do that automatically. Severity, shape and location of these penalties is your choice.