EN-1 A = N-1E[z(k)jo, 0],	 B = E EN-1 N-1C = E[z(k)z(k)Tr lo, 0], D = E : E[z(k)jo, 0],	 B = E[z(k)z(k)Tr lo, 0], k=0 Eq. 12.8 is another third-order nonlinear algebraic equation (in 4 , and t) of the form: N4.T1.4.t — — N4,11"t — N4,t + 4,T 3 + 4, A + Nt — B = 0.	 (12.12)

doi:10.1201/9781482276237-68

Chapter

EN-1 A = N-1E[z(k)jo, 0], B = E EN-1 N-1C = E[z(k)z(k)Tr lo, 0], D = E

ABSTRACT

Denoting by i(kIk) the EKF state estimate and by i(k+llk) the one-step EKF state prediction, the prediction equation for the special structure of our model has the form

1(k + llk) = + (I — (1,)t.

The physical interpretation of Eq. 12.6.3 applied to our speech model is that the one-step EKF state predictor based on the current EKF state estimate will always move towards the target vector t for a given system matrix .1.. Such desirable dynamics comes directly from state equation Eq. 12.4, and it is, in fact, in exactly the same form as the noise-free model state equation.