- You can use parameter covariance matrix to derive an upper bound on individual parameter uncertainties. An example can be sqrt(getcov(model)). Such bounds can be quite conservative since they ignore cross-covariance.
- You can create an LTI array representing a random sample set from the identified model using the "rsample" function. You can then do your control design against this LTI array.
- You can create frequency response uncertainty bounds (see bode, freqresp functions in System Identification Toolbox). You can then use ultidyn functionality from Robust Control toolbox to create uncertain LTI dynamics which can then be used in global H-inf approaches for control design.
How do I fit a transfer function from my measured data with an uncertainty using System Identification Toolbox R2019a?
조회 수: 2 (최근 30일)
이전 댓글 표시
MathWorks Support Team
2020년 4월 7일
답변: MathWorks Support Team
2020년 4월 7일
I have two questions in this regard:
1.) Can I use the hinfstruct directly on the datatype idfrd without using tfest first?
2.) I want to fit a transfer function from my measured data with an uncertainty, which I can then process further. How can this be done?
채택된 답변
MathWorks Support Team
2020년 4월 7일
Structured H-inf requires a parametric plant model. So you need to create dynamic model first. Once such a model is identified (say using tfest or ssest), you have a few options:
댓글 수: 0
추가 답변 (0개)
참고 항목
카테고리
Help Center 및 File Exchange에서 Linear Model Identification에 대해 자세히 알아보기
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!