Parameter Estimation issue: can't re use optimized parameters for resimulating the model

2 posts / 0 new
Last post
markdh
Offline
Joined: 2018-01-15
Parameter Estimation issue: can't re use optimized parameters for resimulating the model

Hello i'm solving a large parameter estimation problem: i've defined a model and an optimization for it that select which parameters to optimize( setting their free attribute) and define an objective function that is the sum of least squared error between some state variable (that depends on the parameters) at a  given time and measurement data points. The problem is that i get different result for state variables in the optimization result, from the one obtained resimulating the model with parameters setted with their optimal values, extracted from the optimization result file.

My model is quite complicated: it has 35 state variables and more than 70 parameters to be optimized plus some algebraic values and some function calls in the  state variables( i can post it if needed).

I'm trying to understand if it is a technical problem( maybe something in my model is not completely supported) or if it is a modelling one(maybe the solver is optimizing some other variable other than the parameters, even if inspecting the results it didn't seem like this to me).

Thank you

andrewsymond
Offline
Joined: 2018-02-12
THANKS

The issue is that I get a distinctive outcome for state factors in the advancement result, Do my Coursework for me from the one acquired simulating the model with parameters set with their ideal esteems removed from the streamlining result document. 

Login or register to post comments