1 / 33

Model And Filter

Model And Filter. Name: 韩 瑜 Email : hanyu@sjtu.edu.cn. Background. Black Box Signal evolution Noise Input Output. Model. Why choose Differential Equations. Realization Circuit (time-delay) Analysis ODE Fourier. Approximation in theory. Curve Fitting

isaac-cox
Télécharger la présentation

Model And Filter

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Model And Filter Name: 韩 瑜 Email:hanyu@sjtu.edu.cn

  2. Background • Black Box • Signal evolution • Noise • Input • Output

  3. Model

  4. Why choose Differential Equations • Realization • Circuit (time-delay) • Analysis • ODE • Fourier

  5. Approximation in theory • Curve Fitting Find original formula of system • Function Approximation Simplify system (has known system precisely)

  6. Curve Fitting Usually Solve this problem

  7. For instance

  8. Least square error

  9. Classic Filter Theory Review • Signal Process

  10. Modern Filter • Wiener Filter • Karlman Filter • Adaptive Filter • Wavelet Filter

  11. Method in H2 Space • LMS • Fitting and Approximation are same (Mention that approximation use Hilbert polynomial)

  12. Defect • It’s difficult to compute by tradition matrix algorithm in actual fact • Some information are same between each generation, but not used, and do same work again

  13. Kalman’s method • Model • Iteration equation

  14. Variant dimension

  15. Kalman Filter

  16. How to understand P(t|t)&P(t+1|t) Many t and t+1, like: X(t+1|t), X(t|t) P(t+1|t), P(t+1|t) How to understand these?

  17. Relation with Kalman

  18. Recursive method • Kalman recursive method • RLS recursive method

  19. Adaptive Filter • Back to model • if model is change Equal to dimension change or Dimension increasing • Iteration equation

  20. Variant dimension

  21. Identification

  22. Model • Widely Accept • Rewrite • Evolution • Observation

  23. Iteration filter • Add some condition • Iteration • Some equation

  24. Wavelet Filter 1.Wavelet decomposition 2.Coefficients restrain 3.Reconstruction

  25. Wavelet Filter Result

  26. Method in H-inf Space • Fitting And Approximation are not same • Approximation: • 2nd Chebyshev polynomial • Fitting • model

  27. Linear programming For H-inf • linear programming method • Mention:

  28. Future of Filters • Filter of H2 • Research may be useless now for low dimension data • Computer and computation • Filter of H-inf • Linear programming • LMI (?) • H2&Hinf filter (?)

  29. Thank You! Question?

  30. Homework

More Related