系统辨识第四章概要1.ppt

COLLEGE OF ELECTRICAL ENGINEERING MODERN CONTROL SYSTEMS COLLEGE OF ELECTRICAL ENGINEERING Part I Systems and Models CHAPTER 4 Models of Linear Time-Invariant Systems A model of a system is a description of (some of) its properties, suitable for a certain purpose. 4.1 Linear Models and Sets of Linear Models In Chapter 2, a linear time-invariant model is specified by the impulse response {g(k)}1,∞, the spectrum Fv(w)= l|H(eiw)|2 of the additive disturbance, and the probability density function (PDF) of the disturbance e(t): y(t)= G(q) u(t) + H(q) e(t) fe(.), the PDF of e (4.1) with  G(q) = Sk=1,∞ g(k) q-k, H(q) = 1 + Sk=1,∞ h(k) q-k (4.2) Also, the PDF fe is often not specified as a function, but described typically by the first and second moments:  (4.3) Using (3.20), the one-step-ahead prediction for (4.4) can be computed as  y^(t|q) = H-1(q,q) G(q,q) u(t) + [1 - H-1(q,q)] y(t) (4.6) 4.2 A Family of Transfer-Function Models Equation Error Model Structure Linear Regressions Such a model is called a linear regression in statistics, and the vector j(t) is known as the regression vector. ARMAX Model Structure Peudolinear Regressions Output Error Model Structure e u w y Fig. 4.3 The output error model structure With the aid of the vector j(t,q) = [u(t -1) … u(t - nb) - w(t -1, q) … - w(t - nf, q)]T (4.29)we have  y^(t|q) = jT(t,q) q (4.30) Box-Jenkins Model Structure A General Family of Model Structures Table 4.1 Some common Black-box SISO Models as Special Cases of (4.33) A y = B/F * u + C/D * e -------------------------------------