- removing normal operating points from your data
- choosing how the initial condition should be handled (e.g. zero or estimate)
- comparing the estimated model output with the data and seeing if it matches what you are expecting
- ...etc.
Specifying bounds in process model estimation
3 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Is it essential to specify bounds in process model estimation using system identification toolbox? I'm not getting the expected process model from estimation.
0 comentarios
Respuestas (1)
Bill Tubbs
el 20 de Dic. de 2022
Editada: Bill Tubbs
el 20 de Dic. de 2022
Not necessarily. You only need bounds if you think your model is overfitting the data (e.g. due to not enough data points and/or imperfect system excitation) and you know some solutions are wrong due to prior knowledge about the true system.
But before starting to add bounds, make sure you have done everything else correct! Including:
0 comentarios
Ver también
Categorías
Más información sobre Linear Model Identification en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!