I have a very complex nonlinear model (derived from first principle's) of a chemical process with all identifiable parameters. The parameters of the gray box model were identified through a stochastic search based system identification method. But the identified model is not suitable for control and observer design because of its complex nature. As suggested in many literature parametric senility analysis would be very much effective for model complexity reduction. But I am facing difficulties in carrying out parametric sensitivity analysis with the identified model. That's why I am thinking of transforming my model into PNSS form, which would make sensitivity analysis and model reduction studies much simpler.
Thank you in advance for your help.
- Badges
- Science topic
- Similar topics
- Mathematics