Model Identification and Data Analysis

Index

• b
• Bayes  212–23
  • formula  214
  • geometric interpretation of batch formula  220–2
  • geometric interpretation of recursive formula  222–3
  • recursive scalar formula  215–17
  • recursive vector formula  219–20

• c
• canonical decomposition  256
• canonical representation  68
• complexity
  • via cross‐validation  157
  • AIC  160–1
  • FPE  157–60
  • MDL  161–4
• control
  • data‐driven design  205–8
  • GMV  196–204
  • MPC  204–5
  • MV  188–96
  • PI controller  206
  • predictive  187
• convolution 43
• covariance
  • coefficient 7–8
  • function 10

• d
• detectability  257
• Diophantine equation  79, 190
• Durbin‐Levinson  168

• e
• eigenvalue  335–7
• eigenvector  335

• f
• filter
  • all‐pass  67
  • pass‐band 36
  • whitening  70
• forgetting factor  150

• g
• Gauss  85

• h
• Hubble  90

• i
• identification
  • 4SID method  183
  • black box  82
  • LS  107
  • LS vs YW  170–1
  • ML  115
• identification (contd.)
  • model identification  82
  • parameter in given model  281
  • signal identification  82
  • of state‐space models  173
  • subspace methods  173
  • white box  281
• innovation  217–19, 224

• k
• Kalman  209
  • extended Kalman filter  271–3
  • filter  209, 239
  • gain  232
  • life  209–10
  • multi‐step‐ahead predictor  237
  • one‐step‐ahead predictor  223–7
  • steady state  240–65
• Kolmogorov life  61

• l
• least squares  107
• long division 18
• Lyapunov
  • algebraic equation ALE  245
  • difference equation DLE  245

• m
• matrix
  • condition number  353–5
  • Hankel  175–6
  • inversion lemma  141, 340
  • norm  353–5
  • positive definite 6
  • positive semi‐definite 6
  • singular values  180
  • SVD  180, 351–3
  • Sylvester inequality  ...