How can set initial conditions of a state space model obtained by data?

Hello I have a problem with the simulation of a state space model.

The model is from a heat process, and I want to set the condition that it

starts at 25°C (Ambient Temperature) and can never fall below it.

Since I obtained the model from data with system identification, I know nothing

about the nature of the state vector to set the initial conditions and here is the problem.

How can I set the initial conditions in such a way, that I will obtain 25°C at the beginning of

my simulation in simulink?

Best regards

Answer 1

[Rajiv Singh]

Is the temperature an output of the model? If so, what is the corresponding input?

Supposing the input and outputs are known (measurable or their target values known), user can use data2state function to compute initial states corresponding to certain I/O history. Suppose temperature is the output. You only want a certain value (25 deg) at the start of the simulation; you say nothing about the history (times -1, -2, ...) of the I/O signals. Hence there will not be a unique solution for initial states for any model with more than one states. For more reliable answers, provide a history of sample length greater than the number of states.