ABSTRACT
JAGAN: “dk3189_c006” — 2006/3/13 — 11:47 — page 371 — #1
In Chapter 5, the adaptive neural network (NN) control of a class of strict feedback nonlinear discrete-time systems was presented using NN. Although Lyapunov stability analysis was discussed in detail, the analysis was limited to a restrictive class of nonlinear systems in strict feedback form. The dynamics of several industrial systems, for example, spark ignition (SI) engine (Daw et al. 1997) during lean operation and with high exhaust gas recirculation (EGR), can be represented only in nonstrict feedback form. Moreover, the controller design presented in Chapter 5 for strict feedback nonlinear systems is not applicable to nonstrict feedback nonlinear discrete-time systems. Therefore, in this chapter, we initially treat the design of the controller by assuming that the states of the nonlinear discrete-time systems in nonstrict feedback form are available for measurement (He and Jagannathan 2003a) and later relax this assumption (He and Jagannathan 2004) by using an NN observer.
